Skip to main content

Full text of "Karl August Wittfogel - Oriental Despotism"

See other formats

(yriental *^^yespotism 


by i/Carl( fr. VVittfogel 

New Haven and London: YALE UNIVERSITY PRESS 









For permission to quote from the following works the author is very 
grateful: H. Idris Bell, Egypt from Alexander the Great to the Arab 
Conquest, Oxford, Oxford University Press, 1948; J. A. Dubois, Hindu 
Manners, Customs and Ceremonies, tr. and ed. Henry K. Beauchamp, 
Oxford, the Clarendon Press, 1943; Dwight D. Eisenhower, Crusade 
in Europe, Garden City, Long Island, Doubleday, 1948; Kautilya's 
Arthds&stra, tr. R. Shamasastry, 2d ed. Mysore, Wesleyan Mission 
Press, 1923; S. N. Miller, "The Army and the Imperial House," in 
Cambridge Ancient History, Cambridge, England, Cambridge Uni- 
versity Press, 1939; Vincent A. Smith, Oxford History of India, Ox- 
ford, the Clarendon Press, 1928; W. W. Tarn, Hellenistic Civilization, 
London, Edward Arnold, 1927; Alfred M. Tozzer, Landa's Relacion 
de las Cosas de Yucatan, Papers of the Peabody Museum of American 
Archaeology and Ethnology, Cambridge, Harvard University, 1941; 
John A. Wilson, "Proverbs and Precepts: Egyptian Instructions," in 
Ancient Near Eastern Texts, ed. James B. Pritchard, Princeton, 
Princeton University Press, 1950. 


Two aspects of this study of Oriental despotism quickly aroused 
interest: the attempt to establish the peculiarity of a non-Western 
semi-managerial system of despotic power and the interpretation of 
Communist totalitarianism as a total managerial, and much more 
despotic, variant of that system. 

A third aspect has been less commented on, but it is largely respon- 
sible for whatever insights the inquiry achieved: the use, of big struc- 
tured concepts for the purpose of identifying big patterns of societal 
structure and change. 

To be sure, this method is not new. It was employed by Aristotle, 
Machiavelli, and the physiocrats. It produced spectacular results when 
Adam Smith and his successors erected a system of economics that 
considered the minutiae of the workshop and market within the con- 
text of the over-all economic and social order. 

Then followed years of indifference. But today the method is again 
coming to the fore. Comprehensive analytic tools are needed for the 
understanding of our complex national and international industrial 
economy. They are vital for a realistic appraisal of the complex 
operations of the Communist world. Today the economists are clamor- 
ing for a new macro-eConomy. And social scientists in other disciplines 
are as eager to find what may be called macro-analytic methods of 

The macro-analytic revolution is the most promising development 
in our present intellectual crisis. But it will be successful only if we 
face empirical reality in its geo-historical depth, and if we include in 
our arsenal the tested big concepts of our intellectual forebears. 
Efforts to appraise the phenomena of Communist totalitarianism, such 
as collective leadership and autocracy, power economy and subsistence 
economy, self-perpetuation and self-liquidation, will do more harm 
than good if we rely primarily on the experiences of multicentered 
societies and neglect the only major precedent of enduringly success- 
ful total power: Oriental despotism. Efforts to explain the agrarian 
crises in the USSR and Communist China will yield problematic re- 
sults if we view Soviet agriculture in terms of American agriculture 
and Chinese agriculture in terms of Soviet agriculture. Such efforts 
are macro-analytic in intent, but meso-analytic in substance. They 
improperly generalize from a limited, and inadequate, empirical base. 

A genuinely macro-analytic inquirer will carefully utilize the theo- 
retical heritage of his field just as does the engineer who endeavors to 



exhaust the creative possibilities of his craft on the earth, under the 
sea, and in space. A scientist who thinks he must invent all his tools 
anew may well enter the research situation with an empty mind — but 
he will also leave it with an empty mind. Properly applied, the growth 
potential of a reality-tested big concept is enormous. Rooted in past 
experiences and ideas, it has every chance to develop with the new 
empirical data that it is likely to uncover. 

Macro-analytical principles guided me when, in the early thirties, 
I tried to determine the peculiarity of Chinese economics as part of 
a peculiar Chinese (and "Asiatic") society. They guided me when, in 
the early forties, I tried to determine the difference between China's 
conquest dynasties and the typically Chinese dynasties. They guided 
me when I tried to determine the difference between Oriental despot- 
ism, the multicentered societies of the West (and Japan), and Com- 
munist (and Fascist) totalitarianism. These same principles continue 
to guide me in my comparative study of total and totalitarian power 

The present volume reproduces the original text of Oriental 
Despotism with a few additions and corrections from the third Ameri- 
can printing and the German edition. For his work in the preparation 
of the German edition I wish to thank Frits Kool, Amsterdam, with 
whom I discussed many of its problems in an extended correspondence. 

The original study received friendly support from many institu- 
tions and persons. I am profoundly indebted to the Far Eastern and 
Russian Institute of the University of Washington for enabling me 
to engage in the diverse research that constitutes the factual basis of 
the present book. As co-sponsor of the Chinese History Project, New 
York, Columbia University provided facilities of office and library. 
For a number of years the Rockefeller Foundation supported the 
over-all project of which this study is an integral part. Grants given 
by the American Philosophical Society and the Wenner-Gren Founda- 
tion for Anthropological Research made possible the investigation of 
special aspects of Oriental despotism. 

Scholars from various disciplines have encouraged my efforts. With- 
out attempting to list them all, I mention in gratitude Pedro Armillas, 
Pedro Carrasco, Chang Chung-li, Nathan Glazer, Waldemar Gurian, 
Karl Menges, Franz Michael, George P. Murdock, Angel Palerm, 
Julian H. Steward, Donald W. Treadgold, Hellmut Wilhelm, and 
C. K. Yang. I have been privileged to discuss crucial problems with 
two outstanding students of modern totalitarianism: Bertram D. 
Wolfe and the late Peter Meyer. 


In the field of the Muslim and pre-Muslim East I was particularly 
aided in my researches by Gerard Salinger. In the realm of Chinese 
studies I drew upon the knowledge of Chaoying Fang, Lienche Tu 
Fang, Lea Kisselgoff, and Tung-tsu Chu, all of whom were, at the 
time of writing, on the staff of the Chinese History Project. Bertha 
Gruner carefully typed and checked the first draft of an analysis of 
Russian society and the Marxist-Leninist attitude toward Oriental 
despotism, intended originally as a separate publication but eventual- 
ly included in significant part in the present volume. Ruth Ricard 
was indefatigable in preparing the manuscript, which offered many 
problems of form, source material, and bibliography. 

An inquiry into the nature of bureaucratic totalitarianism is bound 
to encounter serious obstacles. Among those who helped in overcom- 
ing them, two persons must be mentioned particularly. George E. 
Taylor, director of the Far Eastern and Russian Institute of the Uni- 
versity of Washington, never wavered in his understanding of my 
endeavors and in his support for what seemed at times beyond hope 
of completion. My wife and closest collaborator, Esther S. Goldfrank, 
shared every step in the struggle for the clarification of basic scientific 
truths and human values. 

It was my belief in these values that put me behind the barbed 
wire of Hitler's concentration camps. My final thoughts go to those 
who, like myself, were passing through that inferno of total terror. 
Among them, some hoped for a great turning of the tables which 
would make them guards and masters where formerly they had been 
inmates and victims. They objected, not to the totalitarian means, but 
to the ends for which they were being used. 

Others responded differently. They asked me, if ever opportunity 
offered, to explain to all who would listen the inhumanity of totalitar- 
ian rule in any form. Over the years and more than I can express, 
these men have inspired my search for a deeper understanding of the 
nature of total power. 


New York } September 1962 




Chapter i: The Natural Setting of Hydraulic Society 

A. Changing Man in Changing Nature 

B. The Historical Place of Hydraulic Society 

C. The Natural Setting 
i. Historical Conditions Being Equal, a Major Natural Difference 

the Possible Cause of Decisive Institutional Differences 

2. Several Natural Factors Essential to Farming ,a 

3. Some Essential Factors Defy Compensating Action; Others Respond 
More Readily 

4. The Specific Qualities of Water ,, 
D. Must the Hydraulic Potential Be Actualized? 15 

1. An Open Historical Situation — but Recognizable Patterns of Re- 
sponse jk 

2. The Recognized Advantages of Irrigation Agriculture 17 

a. If ... , then ... t * 

b. Arid, Semi-arid, and Humid Areas: Hypothetical Patterns of Inter- 
action and Growth jd 


Chapter 2: Hydraulic Economy — a Managerial and Genuinely Political 

Economy 22 

A. Division of Labor in Hydraulic Agriculture 23 
1. Preparatory and Protective Operations Separated from Farming 

Proper 2 a 

a. Large-scale Preparatory Operations (Purpose: Irrigation) 23 

b. Large-scale Protective Operations (Purpose: Flood Control) 23 
2. Cooperation 25 

a. Dimension 25 

b. Integration 26 

c. Leadership 26 

d. Hydraulic Leadership — Political Leadership 27 

B. Heavy Water Works and Heavy Industry 27 

C. Calendar Making and Astronomy — Important Functions of the Hy- 
draulic Regime 29 

D. Further Construction Activities Customary in Hydraulic Societies 30 

1. Nonagrarian Hydraulic Works o Q 

a. Aqueducts and Reservoirs Providing Drinking Water 30 

b. Navigation Canals ^i 

2. Large Nonhydraulic Constructions 34 
a. Huge Defense Structures 34 



b. Roads 37 

c. Palaces, Capital Cities, and Tombs 39 

d. Temples 40 

E. The Masters of Hydraulic Society — Great Builders 42 
1. The Aesthetic Aspect 42 

a. Uneven Conspicuousness 42 

b. The Monumental Style 49 

c. The Institutional Meaning 44 

F. The Bulk of All Large Nonconstructional Industrial Enterprises Man- 
aged also by the Hydraulic Government 45 

1. A Comparative View 45 

2. The Power of the Hydraulic State over Labor Greater Than That 

of Capitalist Enterprises 47 

G. A Genuine and Specific Type of Managerial Regime 48 

Chapter 3: A State Stronger than Society 49 

A. Nongovernmental Forces Competing with the State for Societal Lead- 
ership 49 

B. The Organizational Power of the Hydraulic State 50 

1. The Great Builders of Hydraulic Society — Great Organizers 50 

2. Fundamentals of Effective Organization: Counting and Record 
Keeping 5 o 

3. Organizational and Hydraulic Management 52 

a. The Organizational Task Inherent in Large Constructions, Hy- 
draulic and Otherwise 52 

b. Hydraulic Management 52 

4. The Organization of Quick Locomotion and Intelligence 54 

5. The Organizational Pattern of Warfare in Hydraulic Society 59 

a. Monopolization and Coordination 59 

b. Training and Morale 60 

c. Organization of Supplies 61 

d. Planned Warfare and Military Theory 62 

e. Numbers 63 

f. Percentages 66 

C. The Acquisitive Power of the Hydraulic State 67 

1. Organizational and Bureaucratic Prerequisites 67 

2. Labor on the Public Fields and /or the Land Tax 68 

3. Universality and Weight of the Hydraulic Tax Claim 70 

4. Confiscation 72 

D. Hydraulic Property — Weak Property 78 

1. Four Ways of Weakening Private Property 78 

2. Hydraulic Laws of Inheritance: the Principle 79 

3. The Application 79 

4. The Effect 80 
a. On Regulated Villages 80 


b. On Holders of Small Private Property 80 

c. On Holders of Large Private Property 80 

5. Pertinent Western Developments 81 

a. The Democratic City States of Ancient Greece 81 

b. The United States after the War of Independence 81 

c. A Spectacular Contrast: the Strength of Landed Property in Late 
Feudal and Postfeudal Europe 81 

6. Different Social Forces Opposed to Proprietary Perpetuities 82 

a. Small and Mobile Property 82 

b. The States of Feudal and Postfeudal Europe 83 

c. Hydraulic Absolutism Succeeded Where the States of Occidental 
Feudalism and Absolutism Failed 84 

7. The Organizational Impotence of Hydraulic Property Holders 85 

E. The Hydraulic Regime Attaches to Itself the Country's Dominant Re- 
ligion 87 

1. Sole, Dominant, and Secondary Religions 87 

2. Religious Authority Attached to the Hydraulic State 87 

a. The Hydraulic Regime — Occasionally (quasi-) Hierocratic 87 

b. The Hydraulic Regime — Frequently Theocratic 90 

c. Agrarian Despotism Always Keeps the Dominant Religion Inte- 
grated in Its Power System 96 

d. The Changing Position of the Dominant Priesthood in Hydraulic 
Society 99 

F. Three Functional Aspects, but a Single System of Total Power 100 

Chapter 4: Despotic Power — Total and Not Benevolent 101 

A. Total Power 101 

1. Absence of Effective Constitutional Checks 101 

2. Absence of Effective Societal Checks 102 

a. No Independent Centers of Authority Capable of Checking the 
Power of the Hydraulic Regime J02 

b. The So-called Right of Rebellion 103 

c. Election of the Despot — No Remedy 104 

d. Intragovernmental Influences: Absolutism and Autocracy 106 

3. Laws of Nature and Patterns of Culture — No Effective Checks Either 107 

B. The Beggars' Democracy 108 

1. The Managerial Variant of the Law of Changing Administrative 
Returns 109 

a. Hydraulic Agriculture: the Law of Increasing Administrative Re- 
turns 109 

b. The Law of Balanced Administrative Returns 109 

c. The Law of Diminishing Administrative Returns 1 10 

d. Ideal Curve and Reality of Changing Returns 110 

e. Nonhydraulic Spheres of Political Economy 110 

2. The Power Variant of the Law of Changing Administrative Returns 111 
a. Imperative and Worth-while Efforts m 



b. The Forbidding Cost of Total Social Control in a Semimanagerial 

c. Total Social Control Not Necessary for the Perpetuation of Agro- 
managerial Despotism 112 

3. Sectors of Individual Freedom in Hydraulic Society 113 

a. Limitations of Managerial Control 113 

b. Limitations of Thought Control 114 

4. Groups Enjoying Varying Degrees of Autonomy 115 

a. Less Independence than Frequently Assumed 116 

i. The Family 116 

ii. The Village 117 

iii. The Guilds 120 

iv. Secondary Religions 121 

b. Genuine Elements of Freedom Nevertheless Present 122 

5. Conclusion 124 

a. Politically Irrelevant Freedoms 124 

b. A Beggars' Democracy 125 
C. Hydraulic Despotism — Benevolent Despotism? 126 

1. Total Power — for the Benefit of the People? 126 

2. The Claim and the Reality 126 

a. Operational Necessity Not to Be Confused with Benevolence 126 

b. The Rationality Coefficient of Hydraulic Society 126 

c. Whose Rationality Coefficient? 127 

3. The Rulers' Rationality Optimum Prevails 128 

a. Necessity and Choice in the Policy of the Hydraulic Regime 128 

b. The Rulers' Managerial Optimum 128 

c. The Rulers' Consumptive Optimum 129 

d. The Rulers' Judicial Optimum 131 

4. "Absolute Power Corrupts Absolutely" 133 

5. The Rulers' Publicity Optimum 133 

6. The Two-fold Function of the Benevolence Myth 134 

a. It Stresses the Long-range Interest of the Despotic Regime 134 

b. It Weakens Potential Opposition 135 

c. The Presence of Good Sovereigns and Just Officials Fails to Upset 
the Prevailing Trend 135 

7. Hydraulic Despotism: Benevolent in Form, Oppressive in Content 136 

Chapter 5: Total Terror — Total Submission — Total Loneliness 137 

A. Autonomous Man under Total Power 137 

B. Terror Essential for Maintaining the Rulers' Rationality Optimum 137 

1. The Need 137 

2. Its Official Recognition: "Punishment Is the Kingl" 138 

3. The Morphology of Violence 139 

a. Integrated versus Fragmented Patterns of Violence 139 

b. Controlled versus Uncontrolled Violence 140 

C. The Terror of Hydraulic Despotism 140 
1. Its Physical Aspect 141 


2. Its Psychological Aspect 141 

a. Unpredictability j^j 

b. Lenin: "... power not limited by any laws" 141 

c. Lawless Terror and Terror by Law 142 

3. "Government by Flogging" 143 

a. Terror in Managerial Procedures 143 

b. Terror in Fiscal Procedures 143 

c. Terror in Judicial Procedures 144 

d. Western Correspondences Noteworthy for Their Temporary 
Strength and Their Limitations 145 

4. Varying Configurations of Terror in the Hydraulic World 147 

a. Relatively Lenient Developments 147 

b. Average and Excessive Developments 148 

D. Total Submission 149 

1. Man's Response to the Threat of Total Terror 149 
a. The Postulate of Common Sense and the Virtue of Good Citizen- 
ship: Obedience 149 

2. Preparation for Total Obedience: Disciplinary Education 151 

3. The Great Symbol of Total Submission: Prostration 152 

E. Total Loneliness 154 

1. Loneliness Created by Fear 154 

a. The Ruler: Trust No One! 155 

b. The Official: Eternal Suspicion 155 

c. The Commoner: the Fear of Being Trapped by Involvement 156 

2. The Alienation Potential of Total Power 156 

3. Every-day Adjustments 157 

4. Total Loneliness in the Hour of Doom 157 

Chapter 6: The Core, the Margin, and the Submargin of Hydraulic So- 
cieties 161 

A. Preliminary Stock-taking in the Middle of the Journey 16 1 

1. Some Basic Results 16 1 

2. Three Problems Deserving Further Investigation 161 

3. Problems of Hydraulic Density 162 

B. Hydraulic Core Areas 162 

1. How Continuous Is the Hydraulic System of a Given Hydraulic 
Area? 162 

2. How Great Is the Economic and Political Weight of a Given Hy- 
draulic Economy? 163 

3. How Strong Is the Second Major Element of Hydraulic Operation: 
Flood Control? 165 

4. Compact and Loose Hydraulic Societies 165 

5. The Great Agromanagerial Empires — Usually Loose Hydraulic So- 
cieties 166 

6. Degrees of Hydraulic Density and Degrees of Bureaucratic Density 167 

a. The Principle 167 

b. Changing Bureaucratic Density of a Hydraulic Territory 169 


7. Hydraulically Concerned and Hydraulically Unconcerned Masters 

of Hydraulic Society 169 

8. Periods of Agromanagerial Adjustment, Degeneration, and Restora- 
tion 171 

9. The Staying Power of Deteriorated Agromanagerial Hydraulic So- 
cieties 172 

C. The Margin of the Hydraulic World 173 

1. Varying Operational and Bureaucratic Density Patterns in Marginal 
Areas of the Hydraulic World 173 

2. The Growth of Proprietary Forces 176 

3. The Institutional Staying Power of Marginal Oriental Despotism 177 

a. Bureaucratic Interests Favoring the Reproduction of the Despotic 
Order 177 

b. Late Byzantium: Marasmus, rather than Creative Transformation 178 

c. The Extraordinary Staying Power of Tsarist Bureaucracy 179 

d. Ottoman Turkey 181 

e. Diversified Final Evolutions 181 

4. Marginal Agrarian Despotisms Containing Conspicuous Hydraulic 
Elements 182 

a. The Liao Empire 182 

b. Maya Society 184 

5. "Loose 2" or "Marginal 1"? 188 

6. Fragmenting Patterns of Inheritance and a Government-Dependent 
Dominant Religion 188 

7. Location, Genesis, and Institutional Vulnerability of Marginal 
Agrarian Despotisms igi 

a. Location 

b. Genesis 

c. Institutional Vulnerability 194 

D. The Submarginal Zone of the Hydraulic World 195 

1. The Phenomenon 

2. Cases 

a. Protohistorical Greece 

b. Early Rome 197 

c. Japan ig7 

d. Pre-Mongol (Kievan) Russia 201 

3. Comment 203 

E. Societies Which Cross the Institutional Divide 204 
1. Nonagricultural Peoples adopting and transmitting Power Devices 

of Agrarian Despotism 204 

a. Such Devices Not Necessary for, but Compatible with, Nomadic 

b. The Brittleness of Orientally Despotic Power at the Pastoral 
Fringe of the Hydraulic World 206 

2. Agricultural Civilizations Crossing the Institutional Divide 207 

a. Greece 207 


J 95 



b. Rome 208 
i. The Rise of a Hellenistic Version of Oriental Despotism 208 

ii. The Fall of Agromanagerial Despotism in Western Rome 212 

c. Europe after 476 213 

i. Unsuccessful Attempts to Rule Absolutely 213 

ii. The "Unparalleled" Case of the Domesday Book 213 

d. Spain 214 

i. Oriental Conquest 214 

ii. The Reconquista 216 

e. The Introduction of Oriental Despotism into Russia 219 

F. Structure and Change in the Density Patterns of the Oriental World 225 

1. Structure 225 

a. Density Subtypes of Hydraulic Society 225 

b. Differing Frequencies of Occurrence 226 

c. Decreasing Importance of Hydraulic Economy Proper 226 

2. Capacity for Societal Change 227 

Chapter 7: Patterns of Proprietary Complexity in Hydraulic Society 228 

A. The Human Relation Called "Property" 228 

B. Objects of Property Rights 229 

C. The Potential Scope of Proprietary Rights 229 

D. Three Major Complexity Patterns in Hydraulic Civilizations 230 

1. Simple, Semicomplex, and Complex Patterns of Property 230 

2. Supplementary Remarks 231 

a. "Simple I" and "Simple II" 231 

b. Proprietary Complexity and Hydraulic Density 232 

E. Nonspecific and Specific Aspects in the Proprietary Conditions of 
Tribal Hydraulic Societies 232 

1. Nonspecific Aspects 232 

2. Specific Aspects 233 

3. Simple I . . . 237 

F. Patterns of Property in State-centered Simple Hydraulic Societies 238 

1. Statehood versus Primitive Government 238 

2. Steps in the Professionalizing of Government 239 

a. Chagga Chieftainship and the State of Ancient Hawaii 239 

b. Proprietary Consequences 242 

3. Simple Patterns of Property in Land, Industry, and Commerce 243 

4. Variants of Simple Patterns of Hydraulic Property and Society 245 

a. Hawaii 245 

b. Inca Peru 246 

c. Pharaonic Egypt 250 

d. Ancient China 251 

e. Sumer 253 

5. Origins of Bureaucratic Capitalism 255 

6. The Hydraulic Sponge 256 


G. Semicomplex Patterns of Hydraulic Property and Society 258 

1. Occurrences 258 

a. Pre-Conquest Meso-America 258 

b. India, China, the Near East 260 

c. Byzantium and Russia 260 

2. How Powerful Could the Representatives of Private Mobile and Ac- 
tive Property Become in Semicomplex Hydraulic Societies? 262 

a. Miscellaneous Developments 263 

b. Hindu India 264 

c. Ancient Mesopotamia 267 

d. Conclusions 269 
H. Complex Patterns of Property in Hydraulic Society 270 

1. Hydraulic Landlordism, Past and Present 270 

2. Government-controlled and Private Land in Hydraulic Society 271 

a. Types of Government-controlled Land 271 

i. Government-managed Land 271 

ii. Government-regulated Land 272 

iii. Government-assigned Land 273 

b. Private Land 275 

i. Definitions 275 

ii. Origins 275 

c. Types of Landownership 275 

i. Peasant Landownership 275 

ii. Bureaucratic Landlordism 276 

iii. Other Social Groups 277 

iv. Absentee Landlordism (the General Trend) 277 

v. Absentee Landlordism (Traditional Russia) 278 

vi. Borderline Cases of Regulated and Private Land Tenure 279 

d. The Extent of Private Landownership in Various Subtypes of 
Hydraulic Society 280 

i. Simple Hydraulic Societies 280 

ii. Semicomplex Hydraulic Societies 281 

iii. Complex Patterns of Hydraulic Property and Society 286 

3. How Free Is Private Landed Property in Hydraulic Society? 291 

a. Despotically Imposed versus Democratically Established Restric- 
tions of Private Property 291 

b. Restrictions Imposed upon the Freedom to Enjoy, to Use, to 
Transfer, and to Organize 292 

I. The Effect of Private Property on Hydraulic Society 293 

1. The Perpetuation of Hydraulic Society Depends on the Govern- 
ment's Maintenance of its Property Relations 293 

2. The Growing Complexity of Property and the Growing Complexity 

of Society 294 

3. Small Property Offers a Considerable Economic Incentive, but No 
Political Power 294 

a. Incentives Inherent in Private Possession and Ownership 294 

b. The Beggars' Property 295 


4. Private Commercial Property Politically Inconsequential Even When 
Permitted to Become Large 297 

5. Problems of Wealth within the Governing Class 297 

a. Bureaucratic Hedonism 298 

b. Bureaucratic Landlordism and Capitalism 298 

6. Conclusions Leading to New Questions 300 

a. Hydraulic Property: Revenue Property versus Power Property 300 

b. The Importance — and Limitation — of Private Property in De- 
termining Class Differentiations within Hydraulic Society 300 

Chapter 8: Classes in Hydraulic Society 301 

A. The Need for a New Sociology of Class 301 

B. Class Structure in Hydraulic Society 302 

1. The Key Criterion: Relation to the State Apparatus 302 

2. The Multiple Conditioning of Social Subsections 304 

C. The Rulers 305 

1. The Men of the Apparatus 305 

a. The Basic Vertical Structure 305 

i. The Ruler and the Court 305 

ii. The Ranking Officials 306 

iii. The Underlings 307 

b. Horizontal Developments 307 

i. Satraps 308 

ii. Subordinate Princes, Curacas, Rajas 309 

iii. Gradations of Power in Modern Totalitarian States 310 

2. Subclasses Attached to the Men of the Apparatus 311 

a. Attachment Based on Kinship 311 

i. The Ruling House 311 

ii. The Bureaucratic Gentry 312 

iii. The Relatives of Civil Underlings and Rank-and-File Soldiers 316 

b. Attachment Based on Semi-, Quasi-, or Pre-Official Status 317 

i. Secular Semi-officials (Commercial and Fiscal Agents) 317 
ii. Religious Quasi-officials (Functionaries of the Dominant Reli- 
gion) 318 
iii. Persons Occupying a Pre-official Status (Trainees and Degree- 
holding Candidates for Office) 319 
iv. A Comparative Note (Professional Ideologists in the USSR) 319 

c. Subdivided, but Still an Entity 320 

D. The Ruled 321 

1. Property-based Subsections of Commoners 321 

2. Slaves 322 
E. Modifications of Class Structure That Occur in Conquest Societies 324 

1. Conquest Involving the Formation of Stratified Societies (Primary 
Conquest) 324 

2. Conquest Involving the Further Differentiation of Stratified So- 
cieties (Secondary Conquest) 325 


3. Class Modifications in Hydraulic Conquest Dynasties 326 

a. The Chinese Did Not Always Absorb Their Conquerors 326 

b. Devices for Preserving the Conquerors' Hegemony 326 

c. Duplications of Class 327 

F. Many Social Antagonisms but Little Class Struggle 327 

1. Social Antagonism and Class Struggle 327 

2. Paralysis of Class Struggle by Total Power 328 

G. Antagonism between Members of Different Subsections of Com- 
moners 329 

H. The "People" versus the Men of the Apparatus 331 

I. Social Conflicts within the Ruling Class 334 

1. Ranking Officials versus Underlings 335 

a. Bureaucratic Competition 336 

a. Patterns of Competition Different in Different Societies 336 

b. Bureaucratic Competition in Hydraulic Society 337 

3. Civil versus Military Officials 338 

a. The Autocrat and the Army 338 

b. Civil versus Military Officials 339 

4. The Bureaucratic Activists versus the Bureaucratic Gentry 340 

5. Conflicts between the Autocrat and Other Members of the Ruling 
Class 343 

a. The Autocrat versus His Relatives 343 

i. Blood Relatives 343 

ii. Affinals 344 

b. The Autocrat versus the Ranking Officials 345 

i. Once More the Problem of Autocracy 345 
ii. Human (Social) Relations Expressed through Institutional 

Arrangements 345 

6. Autocratic Methods of Controlling the Bureaucratic Personnel 346 

a. The Ruler's Control over a Hereditary Officialdom (a Bureau- 
cratic Nobility) 346 

b. Autocratic Means of Weakening or Destroying the Self-perpetuat- 
ing Quality of the Ranking Officials 346 

i. Priests 346 

ii. Commoners (General Observations) 347 

iii. Commoners: Social Effects and Limitations of the Chinese 

Examination System 347 

iv. Eunuchs: the Principle 354 

v. Eunuchs: a Few Historical Facts 356 

vi. The Despot's Personal Agency No Incipient Party 358 

vii. The Tribal Nobles of Conquest Dynasties 359 

viii. Slaves 360 

7. "Regular" Officials, Control Groups, and the People 362 
J. Social Promotion 363 

1. Reservoirs and Mainsprings of Social Promotion 363 

2. Criteria for Social Promotion (Aptitudes "plus" . . . ) 364 


3. Social Promotion on a Slave Plantation 364 

K. The Total Ruling Class — a Monopoly Bureaucracy 365 

1. The Ruling Class of Hydraulic Society and the Upper Classes in 

Other Stratified Societies 365 

a. Authoritarian Bodies Do Not Necessarily Exert Total Power 366 

3. Monopoly versus Competition in Societal Leadership 366 

4. Monopoly of Societal Leadership Appears in Oriental Despotism as 
Monopoly of Bureaucratic Organization ("Monopoly Bureaucracy") 367 

Chapter 9: The Rise and Fall of the Theory of the Asiatic Mode of Pro- 
duction 369 

A. Old and New Constructs of a Unilinear Development Disregard Hy- 
draulic Society 370 

1. 19th-century Unilinealists 370 

2. Negative Criticisms 370 

3. A Theoretical Vacuum 371 

4. The Spread of a "Marxist-Leninist" Neo-unilinealism 371 

5. The Need for a Reexamination of Marx', Engels', and Lenin's Views 

on the "Asiatic System" and Oriental Despotism 372 

B. Marx, Engels, and Lenin Accept the Asiatic Concept 372 

1. Marx Follows His Classical Predecessors with Regard to the Institu- 
tional Structure and the Developmental Position of the Orient 372 

2. Marx' Asiatic Interpretation of India, China, and Post-Mongol Rus- 
sia 374 

a. India ("Asiatic Society" . . . ) 374 

b. China (". . . Asiatic Production" and Private Peasant Landhold- 

in g) 374 

c. Russia ("Oriental despotism" . . . Perpetuated) 375 

3. Marx Warns against Confusing the State-controlled Agrarian Order 

of Asia with Slavery or Serfdom 376 

4. "General Slavery" 377 

5. For Many Years Lenin Also Upheld the Asiatic Concept 377 

a. "Asiatic Despotism," a Totality of Traits "with Special Economic, 
Political, and Sociological Characteristics" 378 

b. Lenin Elaborates Marx' Semi-Asiatic Interpretation of Tsarist 
Russia 379 

c. Lenin Holds the Term "Feudal" Unsuited to Traditional Russia 379 

C. Retreat from Truth 380 

1. Marx 380 

a. Marx "Mystifies" the Character of the Ruling Class 380 

b. Further Retrogressions 381 

2. Engels 382 

a. Asiatic Society — Yes! (Engels' Basic Attitude) 382 

b. Asiatic Society — Yes and Nol (The Anti-Duhring) 383 

c. Asiatic Society — Nol (The Origin of the Family, Private Property, 
and the State) 384 


d. Retrogressive Trends in a Supposedly Progressive Position 386 
i. Marx Defends Scientific Objectivity against All Extraneous 

Considerations 386 

ii. Marx' and Engels' "Sin against Science" 387 

iii. From Progressive to Reactionary Utopianism 388 

3. Lenin 389 

a. Lenin Further Cripples Marx' Crippled Version of the Asiatic 
Concept 389 

i. Consistent Disregard of the Managerial Aspect of Oriental 

Despotism 389 

ii. A Confused Presentation of Russia's Ruling Class 390 

b. A Power-Strategist's Treatment of Truth 391 

c. The Threat of the Asiatic Restoration (1906-07) 391 

d. Further Oscillations (1907-14) 394 

e. Full Retreat (1916-19) 395 

i. Lenin's Imperialism (1916) 395 

ii. State and Revolution (1917) 396 

iii. Lenin's Lecture on the State (1919) 397 

f. Lenin's Last Period: the Specter of the Aziatchina Reemerges 398 

4. Stalin 401 

a. The Old Guard Objects 401 

b. A Half-hearted Criticism of the Theory of Oriental Society 402 

i. The Leningrad Discussion (1931) 402 

ii. The Significance of the 1931 Discussion 404 

c. Ideological Twilight 405 

d. Stalin "Edits" Marx 407 

e. Delayed Reaction in the Anglo-Saxon World 408 

f. The Rout of the Notorious Theory of the Asiatic Mode of Pro- 
duction 409 

D. Three Forms of the Blackout of the Theory of the Asiatic Mode of 

Production 411 

Chapter 10: Oriental Society in Transition 413 

A. Basic Concepts of Societal Type and Development 414 

1. Societal Types 414 

a. Essential, Specific, and Nonspecific Elements of Society 414 

b. Pre-industrial Stratified Societies 416 

i. Pastoral Society 416 

ii. Several Types of Ancient Societies 416 

iii. Feudal Society 417 

iv. Unwieldy Hydraulic Society 418 

v. Residual Stratified Pre-industrial Societies 418 

2. Societal Changes 419 

a. Forms 419 

b. Values 420 

B. Hydraulic Society in Transition 421 
1. Four Aspects of the Self-perpetuation of Hydraulic Society 421 

a. The Potential for Institutional and Cultural Growth 421 



b. Stagnation, Epigonism, and Retrogression 422 

c. The Staying-Power of Hydraulic Society 422 

d. Societal Change Dependent on External Influence 429 

2. Recent Patterns of External Influence 422 

a. Patterns of Interrelation 424 

b. The Influencing Side 42 ^ 

c. Institutional Differences in the Target Societies 427 

3. Societal Results 42- 

a. Russia 427 

b. Colonized Hydraulic Countries 400 

c. Semidependent ("Semicolonial") Countries 499 

d. A New Developmental Force Arises: Soviet Communism 43G 

4. Hydraulic Society at the Crossroads 436 

a. The Developmental Issue Underlying the Bolshevik Revolution 436 

b. The USSR — Russia's Asiatic Restoration? 408 

c. Communist China — the Product of a Genuine "Asiatic Restora- 

C. Whither Asia? 

D. Whither Western Society — Whither Mankind? 447 
Notes 451 


General Index 

Index of Authors and Works 




When in the 16th and 17th centuries, in consequence of the com- 
mercial and industrial revolution, Europe's trade and power spread 
to the far corners of the earth, a number of keen-minded Western 
travelers and scholars made an intellectual discovery comparable to 
the great geographical exploits of the period. Contemplating the 
civilizations of the Near East, India, and China, they found signifi- 
cant in all of them a combination of institutional features which 
existed neither in classical antiquity nor in medieval and modern 
Europe. The classical economists eventually conceptualized this dis- 
covery by speaking of a specific "Oriental" or "Asiatic" society. 

The common substance in the various Oriental societies appeared 
most conspicuously in the despotic strength of their political au- 
thority. Of course, tyrannical governments were not unknown in 
Europe: the rise of the capitalist order coincided with the rise of 
absolutist states. But critical observers saw that Eastern absolutism 
was definitely more comprehensive and more oppressive than its 
Western counterpart. To them "Oriental" despotism presented the 
harshest form of total power. 

Students of government, such as Montesquieu, were primarily 
concerned with the distressing personal effects of Oriental despotism, 
students of economy with its managerial and proprietary range. The 
classical economists particularly were impressed by the large water 
works maintained for purposes of irrigation and communication. 
And they noted that virtually everywhere in the Orient the gov- 
ernment was the biggest landowner. 1 

These were extraordinary insights. They were, in fact, the starting 
point for a systematic and comparative study of total power. But 
no such study was undertaken. Why? Viewed alone, the social scien- 
tists' withdrawal from the problem of Oriental despotism is puzzling. 
But it is readily understandable when we consider the changes that 
occurred in the 1 9th century in the general circumstances of Western 
life. Absolutism prevailed in Europe when Bernier described his 
experiences in the Near East and Mogul India and when Mon- 
tesquieu wrote The Spirit of the Laxvs. But by the middle of the 1 9th 
century representative governments were established in almost all 
industrially advanced countries. It was then that social science turned 
to what seemed to be more pressing problems. 



Fortunate age. Fortunate, despite the sufferings that an expand- 
ing industrial order imposed on masses of underprivileged men and 
women. Appalled by their lot, John Stuart Mill claimed in 1852 
that "the restraints of Communism would be freedom in comparison 
with the present situation of the majority of the human race." 2 
But he also declared that the modern property-based system of 
industry, outgrowing its dismal childhood, might well satisfy man's 
needs without grinding him down into "a tame uniformity of 
thoughts, feelings, and actions." 3 

Fortunate age. Its ever-critical children could combat the frag- 
mented despotism of privilege and power, because they did not live 
under a system of "general slavery."* Indeed they were so far re- 
moved from the image of absolutist power that they felt no urge to 
study its substance. Some, such as Max Weber, did examine illumi- 
natingly, if not too systematically, certain aspects of Oriental state- 
craft and bureaucracy. But by and large, what Bury said at the close 
of the period of liberalism was true: little effort was made to de- 
termine the peculiarities of absolutism through detailed comparative 

Fortunate age. Optimistic age. It confidently expected the rising 
sun of civilization to dispel the last vestiges of despotism that be- 
clouded the path of progress. 


But the high noon has failed to fulfill the promises of the dawn. 
Political and social earthquakes more terrifying than any that 
previously shook the homelands of modern science make it painfully 
clear that what has been won so far is neither safe nor certain. Total 
power, far from meekly withering away, is spreading like a virulent 
and aggressive disease. It is this condition that recalls man's previous 
experience with extreme forms of despotic rule. It is this condition 
that suggests a new and deepened analysis of Oriental — or as I now 
prefer to call it, hydraulic — society. 


For three decades I studied the institutional settings of Oriental 
despotism; and for a considerable part of this time I was content 
to designate it "Oriental society." But the more my research ad- 

a. Marx (1939: 395) applied this term to Oriental despotism without realizing that 
more comprehensive forms of state slavery might emerge under conditions of industry. 


vanced, the more I felt the need for a new nomenclature. Distinguish- 
ing as I do between a farming economy that involves small-scale 
irrigation (hydroagriculture) and one that involves large-scale and 
government-managed works of irrigation and flood control (hydrau- 
lic agriculture), I came to believe that the designations "hydraulic 
society" and "hydraulic civilization" express more appropriately 
than the traditional terms the peculiarities of the order under dis- 
cussion. The new nomenclature, which stresses human action rather 
than geography, facilitates comparison with "industrial society" and 
"feudal society." And it permits us, without circumstantial reasoning, 
to include in our investigation the higher agrarian civilizations of 
pre-Spanish America as well as certain hydraulic parallels in East 
Africa and the Pacific areas, especially in Hawaii. By underlining 
the prominent role of the government, the term "hydraulic," as I de- 
fine it, draws attention to the agromanagerial and agrobureaucratic 
character of these civilizations. 

The present inquiry goes considerably beyond the findings of the 
early students of Oriental society. In the following pages I endeavor 
to describe systematically man's hydraulic response to arid, semi- 
arid, and particular humid environments. I also indicate how the 
major aspects of hydraulic society interlock in a vigorously function- 
ing institutional going concern. 

This going concern constitutes a geo-institutional nexus which 
resembles industrial society in that a limited core area decisively 
affects conditions in large interstitial and peripheral areas. In many 
cases these marginal areas are politically connected with hydraulic 
core areas; but they also exist independently. Manifestly, the 
organizational and acquisitive institutions of the agrodespotic state 
can spread without the hydraulic institutions which, to judge from 
the available data, account for the genesis of all historically significant 
zones of agrarian despotism. An understanding of the relations 
between the core and the margin of hydraulic society — a phe- 
nomenon barely noted by the pioneer analysts — is crucially im- 
portant for an understanding of Western Rome, later Byzantium, 
Maya civilization, and post-Mongol (Tsarist) Russia. 

In the matter of private property the early institutionalists were 
satisfied to indicate that the Oriental state controlled the strategic 
means of production, and most importantly the cultivable land. The 
real situation is much more complicated and, from the standpoint 
of societal leadership, much more disturbing. History shows that in 


many hydraulic societies there existed very considerable active 
(productive) private property; but it also shows that this development 
did not threaten the despotic regimes, since the property holders, 
as property holders, were kept disorganized and politically impotent. 

Obviously, too much has been said about private property gen- 
erally and too little about strong and weak property and about the 
conditions which promote these forms. The analysis of the varieties 
of private property in hydraulic society determines the limitations 
of nonbureaucratic (and of bureaucratic) private property under 
Oriental despotism. Its results contradict the belief that practically 
any form of avowedly benevolent state planning is preferable to the 
predominance of private property, a condition which modern socio- 
logical folklore deems most abhorrent. 

And then there is the problem of class. Richard Jones and John 
Stuart Mill indicated that in Oriental society the officials enjoyed 
advantages of income which in the West accrued to the private 
owners of land and capital. Jones and Mill expressed a significant 
truth. But they did so only in passing and without stating clearly 
that under agrodespotic conditions the managerial bureaucracy was 
the ruling class. They therefore did not challenge the widely accepted 
concept of class which takes as its main criterion diversities in (active) 
private property. 

The present inquiry analyzes the patterns of class in a society whose 
leaders are the holders of despotic state power and not private owners 
and entrepreneurs. This procedure, in addition to modifying the 
notion of what constitutes a ruling class, leads to a new evaluation of 
such phenomena as landlordism, capitalism, gentry, and guild. It 
explains why, in hydraulic society, there exists a bureaucratic land- 
lordism, a bureaucratic capitalism, and a bureaucratic gentry. It 
explains why in such a society the professional organizations, al- 
though sharing certain features with the guilds of Medieval Europe, 
were societally quite unlike them. It also explains why in such a 
society supreme autocratic leadership is the rule. 6 While the law of 
diminishing administrative returns determines the lower limit of the 
bureaucratic pyramid, the cumulative tendency of unchecked power • 
determines the character of its top. 


The proponent of new scientific ideas unavoidably discards old 
ideas. Almost as unavoidably he will be criticized by those who de- 
fend the old position. Not infrequently such a controversy throws 
new light on the entire issue. This has certainly been the case with 
the theory of Oriental (or hydraulic) society. 


The reader will not be surprised to learn that this theory has 
aroused the passionate hostility of the new total managerial bureauc- 
racy that, in the name of Communism, today controls a large part 
of the world's population. The Soviet ideologists, who in 1931 de- 
clared the concept of Oriental society and a "functional" ruling 
bureaucracy politically impermissible, no matter what the "pure 
truth" might be, 7 cynically admitted that their objections were in- 
spired by political interests and not by scientific considerations. In 
1950 the leaders of Soviet Oriental studies designated as their most 
important accomplishment "the rout of the notorious theory of the 
'Asiatic mode of production.' " 8 

The reference to the "Asiatic mode of production" is indicative of 
the kinds of difficulties that confront the Communist attack on the 
theory of Oriental society. To understand them, it must be re- 
membered that Marx accepted many values of the Western world, 
whose modern private-property-based institutions he wished to see 
destroyed. In contrast to the Soviet conception of partisanship in art 
and science, Marx rejected as "shabby" and "a sin against science" 
any method that subordinated scientific objectivity to an outside 
interest, that of the workers included. 9 And following Richard Jones 
and John Stuart Mill, he began, in the early 1850's, to use the 
concept of a specific Asiatic or Oriental society. Stressing particularly 
the Asiatic system of economy, which he designated as the "Asiatic 
mode of production," Marx upheld the "Asiatic" concept until his 
death, that is, for the greater part of his adult life. Engels, despite 
some temporary inconsistencies, also upheld to the end Marx' version 
of the Asiatic concept. Neither Marx nor Engels clearly defined the 
phenomenon of a marginal Oriental society; but from 1853 on, they 
both emphasized the "semi-Asiatic" quality of Tsarist society and the 
Orientally despotic character of its government. 

Lenin spoke approvingly of Marx' concept of a specific Asiatic 
mode of production, first in 1894 and last in 1914. Following Marx 
and Engels, he recognized the significance of "Asiatic" institutions for 
Tsarist Russia, whose society he viewed as "semi-Asiatic" and whose 
government he considered to be despotic. 10 


I was unaware of the political implications of a comparative study 
of total power when in the winter of 1922-23 and under the in- 
fluence of Max Weber I began to investigate the peculiarities of 
hydraulic society and statecraft. I was unaware of it when, in 1924 
and now with reference to Marx as well as Weber, I pointed to 
"Asiatic" society "■ as dominated by a bureaucratically despotic 


state. 12 I was unaware of having drawn conclusions from Marx' ver- 
sion of the Asiatic concept, which Marx himself had avoided, when in 
1926 and employing Marx' own socio-economic criteria, I wrote that 
Chinese developments in the second half of the first millennium b.c. 
made "the administrative officialdom — headed by the absolutist em- 
peror — the ruling class" 13 and that this ruling class, in China as in 
Egypt and India, was a "mighty hydraulic [Wasserbau] bureauc- 
racy." 14 I elaborated this thesis in 1926, 15 1927, 18 1929," and 1931, 18 
impressed by Marx' insistence on an unbiased pursuit of truth. 6 In 
1932, a Soviet critic of my Wirtschaft und Gesellschaft Chinas de- 
nounced my belief in the objectivity of science. 19 It was at this time 
that the Soviet publishers ceased to print my analyses of Asiatic 
society in general and of Chinese society in particular. 

In the 1930's I gradually abandoned the hope that in the USSR 
the nationalization of all major means of production might initiate 
popular control over the government and the rise of a classless 
society. Deepened understanding of the character of Soviet society 
paved the way to further insights into the structure and ideology of 
bureaucratic despotism. Re-examination of the Marxist-Leninist 
view of Oriental society made it clear that Marx, far from originating 
the "Asiatic" concept, had found it ready-made in the writings of the 
classical economists. I further realized that although Marx accepted 
the classical view in many important essentials, he failed to draw a 
conclusion, which from the standpoint of his own theory seemed 
inescapable — namely, that under the conditions of the Asiatic mode 
of production the agromanagerial bureaucracy constituted the ruling 

Lenin's ambivalence toward the "Asiatic system" is perhaps even 
more revealing. In 1906-07 Lenin admitted that the next Russian 
revolution, instead of initiating a socialist society, might lead to an 

b. I cited Marx' statements on this point in 1927 (Wittfogel, 1927: 296) and again in 
1929 (ibid., 1929a: 581 and n. 60; see also 585). 

c. My article, "Geopolitik, geographischer Materialismus und Marxismus," which 
argued the importance of the natural factor for societal growth in general and for 
Asiatic society in particular (see Wittfogel, 1929: 725-8) was published in Unter dem 
Banner des Marxismus without editorial comment, whereas in the Russian version of 
the same journal (Pod znamenem marxizma, 1929, Nos. 2/3, 6, 7/8) the editor indicated 
his disagreement with some of the author's views. In 1930, the journal refused to 
publish the continuation of my article, which carried farther the analysis of the natural 
foundations of Asiatic society (see Wittfogel, 1932: 593 ff., 597-608). For corrections of 
certain of my early views on the man-nature relationship see below, Chap. 1; cf. Chap. 
9). My book Wirtschaft und Gesellschaft Chinas was translated into Russian, and the 
typewritten translation was circulated among a number of Soviet experts, who were 
asked to write a critical introduction. To my knowledge, such an introduction was 
never written. The translation was never published. 


"Asiatic restoration." But when World War I opened up new pos- 
sibilities for a revolutionary seizure of power, he completely dropped 
the Asiatic concept, which, with oscillations, he had upheld for 
twenty years. By discussing Marx' views of the state without re- 
producing Marx' ideas of the Asiatic state and the Oriental despotism 
of Tsarist Russia, Lenin wrote what probably is the most dishonest 
book of his political career: State and Revolution. The gradual 
rejection of the Asiatic concept in the USSR, which in 1938 was 
climaxed by Stalin's re-editing of Marx' outstanding reference to the 
Asiatic mode of production, logically followed Lenin's abandonment 
of the Asiatic concept on the eve of the Bolshevik revolution. 


The campaign against the Asiatic concept shows the master minds 
of the Communist camp unable to bolster their rejection with ra- 
tional arguments. This in turn explains the oblique and primarily 
negative methods with which the friends of Communist totalitarian- 
ism in the non-Communist world oppose the outlawed concept. To 
the uninitiated these methods, which use distortion and de-emphasis 
rather than open discussion, are confusing. To the initiated they 
disclose once more the scientific weakness of the most powerful attack 
against the theory of Oriental (hydraulic) society. 

The picture of hydraulic society given in this inquiry implies 
definite concepts of societal type and development. No doubt there 
is structure and cohesion in man's personal history. All individuals 
base their behavior on the conviction that the regularities of yester- 
day are necessarily linked to the regularities of today and tomorrow. 
And there is structure and cohesion in the history of mankind. In- 
dividuals and groups of individuals like to speak of institutional 
units which they see operating in the present and which they expect 
to operate, or to change recognizably, in the future. Agnostic with- 
drawal from the problem of development therefore ceases to be 
plausible as soon as it is clearly defined. 

However, the absurdity of developmental agnosticism provides no 
excuse for a scheme of historical change that insists on a unilinear, 
irresistible, and necessarily progressive development of society. Marx' 
and Engels' acceptance of Asiatic society as a separate and stationary 
conformation shows the doctrinal insincerity of those who, in the 
name of Marx, peddle the unilinear construct. And the comparative 
study of societal conformations demonstrates the empirical un- 


tenability of their position. Such a study brings to light a complex 
sociohistorical pattern, which includes stagnation as well as develop- 
ment and diversive change and regression as well as progress. By 
revealing the opportunities, and the pitfalls, of open historical situa- 
tions, this concept assigns to man a profound moral responsibility, 
for which the unilinear scheme, with its ultimate fatalism, has no 


Congruent with the arguments given above, I have started my 
inquiry with the societal order of which agromanagerial despotism is 
a part; and I have stressed the peculiarity of this order by calling it 
"hydraulic society." But I have no hesitancy in employing the tradi- 
tional designations "Oriental society" and "Asiatic society" as syno- 
nyms for "hydraulic society" and "agromanagerial society"; and 
while using the terms "hydraulic," "agrobureaucratic," and "Oriental 
despotism" interchangeably, I have given preference to the older 
formulation, "Oriental despotism" in my title, partly to emphasize 
the historical depth of my central concept and partly because the 
majority of all great hydraulic civilizations existed in what is cus- 
tomarily called the Orient. Originally I had planned to publish 
this study under the title Oriental Society. 

The preservation of the old nomenclature stands us in good stead 
when we examine recent developments. For while there are some 
traces of hydraulic society left in certain regions of Latin America, 
the heritage of the old order is still very conspicuous in many coun- 
tries of the Orient proper. The problem of hydraulic society in 
transition is therefore primarily the problem of this area. 

Under what influences and in what ways are the people of the East 
throwing off the conditions of hydraulic society which they main- 
tained for millennia? The significance of this question becomes 
fully apparent only when we understand that Oriental despotism 
atomized those nonbureaucratic groups and strata which, in feudal 
Europe and Japan, spearheaded the rise of a commercial and in- 
dustrial society. Nowhere, it seems, did hydraulic society, without 
outside aid, make a similar advance. It was for this reason that Marx 
called Asiatic society stationary and expected British rule in India 
to accomplish "the only social revolution ever heard of in Asia" by 
establishing there a property-based non-Asiatic society. 20 

Subsequent events indicate that Marx seriously overrated the 
transformative strength of capitalist economy. To be sure, Western 
rule in India and other Oriental countries provided new possibilities 


for a nontotalitarian development; but at the end of the era of 
Western colonialism and despite the introduction of parliamentary 
governments of various kinds, the political leaders of the Orient are 
still greatly attracted by a bureaucratic-managerial policy which 
keeps the state supremely strong and the nonbureaucratic and private 
sector of society supremely weak. 


In this context, certain aspects of Russia's recent development 
deserve the most careful scrutiny. The marginally Oriental civiliza- 
tion of Tsarist Russia was greatly influenced by the West, though 
Russia did not become a Western colony or semi-colony. Russia's 
Westernization radically changed the country's political and eco- 
nomic climate, and in the spring of 1917 its antitotalitarian forces 
had a genuine opportunity to accomplish the anti-Asiatic social revo- 
lution which Marx, in 1853, had envisaged for India. But in the fall 
of 1917 these antitotalitarian forces were defeated by the Bolshevik 
champions of a new totalitarian order. They were defeated because 
they failed to utilize the democratic potential in a historical situation 
that was temporarily open. From the standpoint of individual free- 
dom and social justice, 1917 is probably the most fateful year in 
modern history. 

The intellectual and political leaders of non-Communist Asia, who 
profess to believe in democracy and who in their majority speak 
deferentially of Marx, will fulfill their historical responsibility only 
if they face the despotic heritage of the Oriental world not less but 
more clearly than did Marx. In the light of the Russian experience 
of 1917 they should be willing to consider the issue of an "Asiatic 
restoration" not only in relation to Russia but also to present-day 


The masters of the modern totalitarian superstate build big and 
integrated institutions, which, they say, we cannot emulate. And they 
display big and integrated ideas, which, they say, we cannot match. 
They are right in one respect. We do not maintain totalitarian sys- 
tems of integrated power and ideology. Favorable constellations of 
historical events have permitted us to avoid these monstrous de- 
velopments that paralyze the search for scientific truth and social 
improvement. But our opponents are wrong when they hold us in- 
capable of voluntary association because we reject the disciplines of 
general (state) slavery. They are wrong when they hold us incapable 


of producing big and structured ideas because we reject state-imposed 

Political freedom is not identical with the absence of organized 
action, though our enemies would be happy if this were so. And 
intellectual freedom is not identical with the absence of integrated 
thought. It is only under the conditions of free discussion that 
comprehensive sets of ideas can be genuinely tested. 

In the recent past, scholars often gave themselves to the study of 
details because they took the broad principles of life and thought 
for granted. Seeing these principles threatened, they today begin to 
recall that the trail blazers of modern thought viewed nature and 
society as integrated orders whose architecture they explored. The 
Newtons, Montesquieus, Adam Smiths, and Darwins provided new 
interpretations of the world that were as spontaneous as they were 
coherent, and as bold as they were competent. 

You cannot fight something with nothing. In a crisis situation, any 
theoretical vacuum, like any power vacuum, invites disaster. There 
is no excuse for letting the enemy have things his way when our side 
possesses infinite reserves of superior strength. There is no excuse 
for letting the totalitarian strategists parade their contrived doctrines 
on ground that is legitimately ours. There is no excuse for letting 
them win the battle of ideas by default. 

Scientific inquiry has its inner laws. But it earns the privilege of 
freedom only when, rooted in the heritage of the past, it alertly faces 
the threats of a conflict-torn present and boldly exhausts the pos- 
sibilities of an open future. 



he natural setting of hydraulic society 


Contrary to the popular belief that nature always remains the 
same — a belief that has led to static theories of environmentalism 
and to their equally static rejections — nature changes profoundly 
whenever man, in response to simple or complex historical causes, 
profoundly changes his technical equipment, his social organization, 
and his world outlook. Man never stops affecting his natural en- 
vironment. He constantly transforms it; and he actualizes a new 
forces whenever his efforts carry him to a new level of operation. 
Whether a new level can be attained at all, or once attained, where 
it will lead, depends first on the institutional order b and second 
on the ultimate target of man's activity: the physical, chemical, and 
biological world accessible to him. Institutional conditions being 
equal, it is the difference in the natural setting that suggests and 
permits — or precludes — the development of new forms of technology, 
subsistence, and social control. 

A waterfall interested primitive man little except as a landmark or 
an object of veneration. When sedentary man developed industry 
on a sophisticated mechanical level, he actualized the motive energy 
of water; and many new enterprises (mills) arose on the banks of 
rushing streams. The discovery of the technical potential inherent 
in coal made man geology conscious as never before, and the water 
mill became a romantic survival in the revolutionized industrial 
landscape dominated by the steam engine. 

a. For the terms "transformation" and "actualization," as used here, see Wittfogel, 
1932: 482. 

b. This formulation differs from my earlier concept of the relation between man 
and nature (Wittfogel, 1932: 483 ft., 712 ft.) in its emphasis on the primary importance 
of institutional (and cultural) factors. From this premise follows the recognition of 
man's freedom to make a genuine choice in historically open situations, a point 
developed in the later part of the present chapter. Except for these corrections — which 
are essential also for my criticism of certain ideas of Marx that I had previously 
accepted — 1 am upholding the substance of my earlier views (see Wittfogel, 1931: 21 ff.; 
ibid., 1932: 486 ff.). 


In recent years man has uncovered the productive energies of 
electricity. Again he is turning his attention to falling water. But 
even when the engineer of the 20th century erects his power plant 
on the very spot that previously supported a textile mill, he actualizes 
new forces in the old setting. Nature acquires a new function; and 
gradually it also assumes a new appearance. 


What is true for the industrial scene is equally true for the agri- 
cultural landscape. The hydraulic potential of the earth's water- 
deficient regions is actualized only under specific historical cir- 
cumstances. Primitive man has known water-deficient regions since 
time immemorial; but while he depended on gathering, hunting, and 
fishing, he had little need for planned water control. Only after he 
learned to utilize the reproductive processes of plant life did he begin 
to appreciate the agricultural possibilities of dry areas, which con- 
tained sources of water supply other than on-the-spot rainfall. Only 
then did he begin to manipulate the newly discovered qualities of the 
old setting through small-scale irrigation farming (hydroagriculture) 
and/or large-scale and government-directed farming (hydraulic agri- 
culture). Only then did the opportunity arise for despotic patterns 
of government and society. 

The opportunity, not the necessity. Large enterprises of water 
control will create no hydraulic order, if they are part of a wider 
nonhydraulic nexus. The water works of the Po Plain, of Venice, and 
of the Netherlands modified regional conditions; but neither North- 
ern Italy nor Holland developed a hydraulic system of government 
and property. Even the Mormons, who established a flourishing 
hydraulic agriculture in the heart of arid North America, never 
succeeded in completely eliminating the political and cultural in- 
fluence of their wider industrial environment. The history of the 
Latter-Day Saints illustrates both the organizational potential of 
large-scale irrigation and the limitations imposed on the development 
of hydraulic institutions by a dominant Western society. 

Thus, too little or too much water does not necessarily lead to 
governmental water control; nor does governmental water control 
necessarily imply despotic methods of statecraft. It is only above the 
level of an extractive subsistence economy, beyond the influence of 
strong centers of rainfall agriculture, and below the level of a 
property-based industrial civilization that man, reacting specifically 
to the water-deficient landscape, moves toward a specific hydraulic 
order of life. 

CHAPTER 1, C lg 


1. Historical Conditions Being Equal, a Major 
Natural Difference the Possible Cause of 
Decisive Institutional Differences 

Many factors differentiated agrarian life prior to the industrial age, 
but none equaled in institutional significance the stimulating con- 
tradictions offered by arid areas possessing accessible sources of 
water supply other than on-the-spot rainfall. Under the just-defined 
conditions of pre-industriai agriculture, this natural configuration 
decisively affected man's behavior as a provider of food and organizer 
of human relations. If he wanted to cultivate dry but potentially 
fertile lands permanently and rewardingly, he had to secure a re- 
liable flow of moisture. Of all tasks imposed by the natural environ- 
ment, it was the task imposed by a precarious water situation that 
stimulated man to develop hydraulic methods of social control. 

2. Several Natural Factors Essential to Farming 

Water is not the only natural factor essential for successful crop 
raising. Anyone wishing to farm must have at his disposal useful 
plants, an arable soil, adequate humidity, appropriate temperature 
(sufficient sun and a proper growing season), and a suitable lay of the 
land (relief, surface). 

All these elements are equally essential. The lack of any one of 
them destroys the agronomic value of all the others. Cultivation 
remains impossible unless human action can compensate for the 
total deficiency of any essential factor. 

3. Some Essential Factors Defy Compensating 
Action; Others Respond More Readily 

The effectiveness of man's compensating action depends on the 
ease with which a lacking natural factor can be replaced. Some factors 
must be considered constants because, under existing technological 
conditions, they are for all practical purposes beyond man's control. 
Others are more pliable. Man may manipulate or, if necessary, change 

Temperature and surface are the outstanding constant elements 
of the agricultural landscape. This was true for the premachine age; 
and it is still essentially true today. Pre-industrial attempts to change 

a. For similar attempts at defining the natural factors basic to agriculture see CM: 
125; SM: 753; Widtsoe, 1928: 19 ff.; Buck, 1937: 101. 


the temperature of farming areas have, for obvious reasons, met with 
no success; and even such achievements as central heating and air 
conditioning have wrought no major change. Still less has man 
succeeded in altering the cosmic circumstances which ultimately 
determine the temperature of the earth. 

The lay of the land has equally defied human effort. Man has 
made many minor adjustments such as leveling or terracing — most 
frequently, it would seem, in connection with operations of hydro- 
agriculture. But before modern power machines and high explosives 
were invented, the globe's relief remained fundamentally unaltered. 
Even machine-promoted agriculture, like the technically less ad- 
vanced forms of farming, prospers on the even surfaces of lowlands 
and high plateaus or on gently graded slopes and hills, and not in 
rugged mountainous terrain. 

Vegetation and soil do not resist human action to any comparable 
degree. The farmer professionally manipulates plants and soils. 
He may transfer useful plants to regions lacking them, and he 
frequently does so. However, such action is sporadic and temporary; 
it ceases when the limited objective is achieved. In a given agri- 
cultural area the operations of crop breeding are repeated again and 
again; but the plants cover the ground discontinuously, and although 
under certain circumstances farm labor may be coordinated in work 
teams, there is nothing in the nature of the individual plants or plant 
aggregates which necessitates large-scale cooperation as a prerequisite 
for successful cultivation. Before the machine age the greater part of 
all agriculture proceeded most effectively when individual husband- 
men or small groups of husbandmen attended to the crops. 

The second variable factor, soil, follows a similar pattern, with 
special limitations dictated by the relative heaviness of pulverized 
mineral substance. While seeds or plants have frequently been trans- 
ferred to deficient areas, soil has rarely been moved to barren regions. 
No doubt, poor or useless fields have been improved by bringing bet- 
ter soil from a distance. But such action is of little consequence for 
the character of any major farming area. 1 Man's efforts seek primarily 
to adjust the existing soil to the needs of the crops by hoeing, digging, 
or plowing, and on occasion by improving its chemical composition 
through the application of fertilizers. 

Thus soil is susceptible to manipulation, but to a type of manipula- 
tion that requires work groups no larger than are necessary for the 
cultivation of the plants. Even when, under primitive conditions, 
the clearing of the ground and the gathering of the harvest are under- 
taken by large teams, the actual task of tilling the fields is usually 
left to one or a few individuals. 


4. The Specific Qualities of Water 

Compared with all other essential natural prerequisites of agri- 
culture, water is specific. Temperature and surface, because of their 
respective cosmic and geological dimensions, have completely pre- 
cluded or strikingly limited human action throughout the pre- 
industrial era and afterward. In contrast, water is neither too remote 
nor too massive to permit manipulation by man. In this regard it 
resembles two other variables, vegetation and soil. But it differs 
greatly from both in its susceptibility to movement and in the 
techniques required to handle it. 

Water is heavier than most plants. It can nevertheless be much 
more conveniently managed. Unhampered by the cohesiveness of 
solid matter and following the law of gravity, water flows auto- 
matically to the lowest accessible point in its environment. Within 
a given agricultural landscape, water is the natural variable par 

And this is not all. Flowing automatically, water appears unevenly 
in the landscape, gathering either below the surface as ground water, 
or above the surface in separate cavities (holes, ponds, lakes), or 
continuous beds (streams, rivers). Such formations are of minor 
significance in an agricultural area enjoying ample precipitation, 
but they become immensely important in the water-deficient land- 
scape. The human operator who has to handle water deals with a 
substance that is not only more mobile than other agronomic vari- 
ables, but also more bulky. 

This last quality presents special difficulties whenever man tries to 
utilize large agglomerations of moisture; and this he is prone to do 
whenever natural and technological conditions permit. No opera- 
tional necessity compels him to manipulate either soil or plants in 
cooperation with many others. But the bulkiness of all except the 
smallest sources of water supply creates a technical task which is 
solved either by mass labor or not at all. 


1. An Open Historical Situation — but 
Recognizable Patterns of Response 

The stimulating contradiction inherent in a potentially hydraulic 
landscape is manifest. Such a landscape has an insufficient rainfall 
or none at all; but it possesses other accessible sources of water 
supply. If man decides to utilize them, he may transform dry lands 


into fertile fields and gardens. He may, but will he? What makes him 
engage in a venture which involves great effort and which is fraught 
with highly problematic institutional consequences? 

Historical evidence reveals that numerous groups of persons have 
made this decision. Yet it also reveals that many others have failed to 
do so. Over millennia, tribal gatherers, hunters, fishermen, and 
pastoralists inhabited potentially hydraulic regions, often in close 
proximity to irrigation farmers, but few abandoned their traditional 
occupations for a hydroagricultural way of life. 

Manifestly, no irresistible necessity compelled man to utilize the 
new natural opportunities. The situation was open, and the hydro- 
agricultural course was only one of several possible choices. Never- 
theless, man took this course so frequently and in so many separate 
areas that we may assume regularity in evaluation as well as in 

Man pursues recognized advantage. Whenever internal or external 
causes suggest a change in technology, material production, or social 
relations, he compares the merits of the existing situation with the 
advantages — and disadvantages — that may accrue from the con- 
templated change. Special effort is required to attain the new ob- 
jective; and this effort may involve not only increased work and a 
shift from pleasant to unpleasant operations, but also social and 
cultural adjustments, including a more or less serious loss of personal 
and political independence. 

When the sum total of the accruing benefits clearly and con- 
vincingly exceeds the required sacrifices, man is willing to make the 
change; but problematic advantage usually leaves him cool. Here, as 
elsewhere, the human budget is compounded of material and non- 
material items; any attempt to formulate it exclusively in terms of 
smaller or larger quantities of things (goods) will prove unsatis- 
factory. To be sure, the material factor weighs heavily, but its relative 
importance can be reasonably defined only when full recognition is 
given to such other values as personal safety, absence of oppression, 
and time-honored patterns of thought and action. 

Culture historians have made much of the fact that during the 
"recent" epoch of geozoology x clusters of persons adopted agricul- 
ture, either as a supplementary occupation or, and increasingly, as 
their main subsistence economy. No doubt this transition pro- 
foundly affected the fate of mankind; but any reference to the law 
of recognized advantage must take into account the many primitive 
groups that did not turn to crop-raising either during the days of 
incipient agriculture or after the rise of powerful and stratified 
agrarian civilizations. 

The agrarian alternative had a limited — and very diverse — appeal 

CHAPTER 1, D 17 

to nonfarming groups when cultivation was primitive and leadership 
not overly demanding. After the emergence of stratified agricultural 
societies, choice became even more serious. The authority wielded 
by the governments and wealthy landowners of nearby agrarian 
states acted as a deterrent, for under these conditions a shift might 
involve submission to distasteful methods of political and proprietary 
control. Often women, children, and war captives tilled some few 
fields close to a camp site; but the dominant members of the tribe, 
the adult males, stubbornly refused to abandon their hunting, fish- 
ing, or herding activities. The many primitive peoples who endured 
lean years and even long periods of famine without making the 
crucial changeover to agriculture demonstrate the immense attrac- 
tion of nonmaterial values, when increased material security can be 
attained only at the price of political, economic, and cultural sub- 

2. The Recognized Advantages of Irrigation 


The transition to irrigation farming poses the problem of choice in a 
still more complex form. The primary choice — whether or not to 
start hydroagriculture where it had not been known previously — 
was generally, though perhaps not exclusively, made by groups 
familiar with the techniques of primitive rainfall farming. 

The secondary (derivative) choice — whether or not to emulate an 
established irrigation economy — confronts the traditional rainfall 
farmer as well as the nonagricultural tribesman. But the nonagricul- 
turist is much less prepared technically and culturally to make this 
shift; and in both cases decision becomes more precarious when 
acceptance of a materially attractive irrigation economy involves 
reduction to an abjectly low social and political status. 

It is obviously for this reason that a number of communities 
practicing rainfall farming in Southwest China, India, and Meso- 
America as well as many tribal hunters, fishermen, and herders on the 
fringe of the hydroagricultural world failed to make the change. The 
fate of those who rejected the ambivalent opportunity varied greatly; 
but whatever their subsequent fortunes, history offered most of them 
a genuine choice, and man proceeded not as the passive instrument 
of an irresistible and unilinear developmental force but as a dis- 
criminating being, actively participating in shaping his future. 

a. If ... , then . . . 

Irrigation farming always requires more physical effort than rain- 
fall farming performed under comparable conditions. But it requires 


radical social and political adjustments only in a special geohistorical 
setting. Strictly local tasks of digging, damming, and water distribu- 
tion can be performed by a single husbandman, a single family, or a 
small group of neighbors, and in this case no far-reaching organiza- 
tional steps are necessary. Hydroagriculture, farming based on small- 
scale irrigation, increases the food supply, but it does not involve 
the patterns of organization and social control that characterize 
hydraulic agriculture and Oriental despotism. 

These patterns come into being when an experimenting com- 
munity of farmers or protofarmers finds large sources of moisture in 
a dry but potentially fertile area. If irrigation farming depends on 
the effective handling of a major supply of water, the distinctive 
quality of water — its tendency to gather in bulk — becomes in- 
stitutionally decisive. A large quantity of water can be channeled 
and kept within bounds only by the use of mass labor; and this mass 
labor must be coordinated, disciplined, and led. Thus a number of 
farmers eager to conquer arid lowlands and plains are forced to 
invoke the organizational devices which — on the basis of premachine 
technology — offer the one chance of success: they must work in co- 
operation with their fellows and subordinate themselves to a direct- 
ing authority. 

Again history followed no unilinear course dictated by unavoid- 
able necessity. There were recognized alternatives; and those who 
were faced with them were able to make a genuine choice. But what- 
ever their decisions, they were made within a framework that offered 
only a limited number of workable possibilities. 

Thus the changeover to hydraulic agriculture, or its rejection, was 
not without order or direction. The various decisions displayed 
regularities in conditioning and motivation. But the relative equality 
of the original choices did not imply a relative equality in the final 
results. The majority of all hunters, fishermen, and rainfall farmers 
who preserved their traditional way of life were reduced to in- 
significance, if they were not completely annihilated. Some groups, 
practicing a mixed economy with little or no hydroagriculture, were 
strong enough to impose their will on adjacent hydraulic civilizations. 

The herders came into their own at a relatively late time and in 
a special geohistorical setting. Often they maintained themselves 
against all manner of agriculturists, and in a number of instances 
they engaged in sweeping offensives, accomplishing conquests that 
profoundly modified the political and social structure of the subdued 
agrarian civilizations. 

The representatives of rainfall farming made history in certain 
areas of the West, which was uniquely suited to this type of economy. 

CHAPTER 1, D ig 

But the hydraulic agriculturists outgrew and outfought the majority 
of all neighboring peoples wherever local conditions and interna- 
tional circumstances one-sidedly favored an agromanagerial economy 
and statecraft. 

The pioneers of hydraulic agriculture, like the pioneers of rainfall 
farming, were unaware of the ultimate consequences of their choice. 
Pursuing recognized advantage, they initiated an institutional de- 
velopment which led far beyond the starting point. Their heirs and 
successors built colossal political and social structures; but they did 
so at the cost of many of those freedoms which the conservative 
dissenters endeavored and, in part, were able to preserve. 

b. Arid, Semi-arid, and Humid Areas: Hypothetical Patterns 
of Interaction and Growth 

In their pursuit of recognized advantage, rainfall farmers ex- 
perimented with hydroagriculture not only in desert-like areas of 
full aridity and steppe-like areas of semi-aridity, but also in humid 
areas suitable to the cultivation of useful aquatic plants, above all 

The first two types of landscapes, taken together, cover almost 
three-fifths 2 — and all three possibly something like two-thirds — of 
the globe's surface. Within this area each of the three types of 
potentially hydraulic landscapes may have played a specific role, 
particularly in the formative period of a hydraulic economy. In 
a major sector comprising all three types, the semi-arid regions 
are highly suitable to small and gradually growing enterprises of 
water control. The arid regions provide an ultimate testing ground 
for the new techniques. And the semi-arid and humid regions profit 
further from the technical and organizational experience gained in 
man's victory over the desert. 

This may well have been the sequence in the spread of hydraulic 
agriculture in such widely separated areas as ancient Mesopotamia, 
India, and the western zone of South America. A different order of 
development is probable for landscapes that are homogeneously 
arid, and still another for those that are predominantly semi-arid. 

In each case, the presence or absence of adjacent humid regions 
complicated the pattern of growth. In Egypt, gatherers, hunters, and 
fishermen seem to have practiced agriculture as a subsidiary oc- 
cupation on the naturally flooded banks of the Nile long before 
farming became the primary pursuit. In Meso-America ° and in 

a. Some twenty years ago I considered Aztec Mexico, like pre-Tokugawa Japan, a 
feudal society with small-scale irrigation (Wittfogel, 1932: 587 ft). On the basis of a 


China diffusion (from South America and Inner or South Asia 
respectively) cannot be excluded. But such external stimulation 
need not have occurred; if it did, it was effective only because the 
rainfall farmers in the "stimulated" areas were ready to recognize 
the advantages of the new technique. 

growing familiarity with the early sources I came to recognize the hydraulic character 
of the core areas of pre-Spanish Mexico; and the recent work of Mexican archaeologists 
and historians fortifies me in my conclusion (see Armillas, 1948: 109; ibid., 1951: 24 ft.; 
Palerm, 1952: 184 ff.). I quote particularly from a study by Palerm which provides a 
wealth of historical data on irrigation in both pre-Spanish and early Spanish Meso- 

4. The majority of the irrigation systems seem to have been only of local 
importance and did not require large hydraulic undertakings. Nevertheless, im- 
portant works were undertaken in the Valley of Mexico, and irrigation appears 
in concentrated form in the headwaters of the rivers Tula, Lerma and Atlixco, 
and in the contiguous area of Colima- Jalisco. 

5. The largest concentrations and most important works of irrigation coincide, 
generally, with the greatest density of population, with the distribution of the 
most important urban centers, and with the nuclei of political power and military 
expansion [Palerm, 1954: 71]. 

How far back can we trace hydraulic activities in Meso-America? Armillas believes 
that the great cultural advance in the Hohokam civilization of Arizona (a.d. 500-900) 
was probably due to the construction of irrigation canals, a fact which is archaeologi- 
cally established. And since the remains point to relations between Hohokam and 
Meso-America, he believes that "the same factor may underlie the cultural develop- 
ment in certain areas of western Meso-America during this period" (Armillas, 1948: 
107). The Hohokam data tie in with the "classical" period of Meso-American history, 
which, in the Mexican lake area, probably began in the early centuries of the first 
millennium a.d. Armillas' assumption is reinforced by a recent pollen analysis, which 
suggests that aridity increased during the late "archaic" period (Sears, 1951: 59 ff.). 
Palerm has stated that this climatic change may have caused "the emergence or 
extension of irrigation" in Meso-America (1955: 35). 

Increasing aridity could explain the appearance of concentrated populations and 
the spread of monumental building in Meso-America. But what we know about 
climatic conditions in postglacial times warns against overrating the significance of 
Sears' valuable findings. The spread of monumental building in Meso-America during 
the early part of the first millennium a.d. may well have been due to less rain and 
more irrigation; but this does not mean that, prior to the "classical" period, precipita- 
tion was sufficiently regular to make recourse to irrigation unnecessary. In fact, recent 
excavations by A. Palerm and E. Wolf point to the existence of hydraulic activities in 
the Mexican lake area by the middle of the first millennium b.c. 

Other investigations undertaken by these two anthropologists indicate a relatively 
late date for the building of comprehensive water works by the territorial state of 
Texcoco, which, when the Spaniards arrived, was second only to Mexico. Manifestly, 
acceptance of the lateness of this development involves no rejection of the earlier 
occurrence of hydraulic activities in other sections of the lake area. Rather, the data 
suggest that Texcoco moved slowly from marginal to more central hydraulic condi- 
tions. (For the problem of changing hydraulic density, see below, Chap. 6.) 

CHAPTER 1, D 21 

In ancient China the semi-arid North and the rice-growing South 
established noteworthy forms of interaction. The ancient Yangtze 
states developed early and perhaps under the influence of the rice 
culture of Southeast Asia; but it was the semi-arid North which, 
over a long period of time, constituted the dominant center of power 
and cultural advance in Eastern Asia. In India the arid, semi-arid, 
and humid regions of the North became historically prominent 
before the excessively humid area of Bengal. 

These developmental sequences are presented as hypotheses. Their 
validity, or lack of validity, is of no consequence to our analysis of 
societal structure. They are worth noting, in the main, because on 
the basis of our present archaeological and prehistorical knowledge 
they suggest a highly dynamic interplay between the various types 
of landscapes which combine to form the larger areas of hydraulic 


ydraulic economy— a managerial and 
genuinely political economy 

The characteristics of hydraulic economy are many, but three 
are paramount. Hydraulic agriculture involves a specific type of divi- 
sion of labor, It intensifies cultivation. And it necessitates coopera- 
tion on a large scale. The third characteristic has been described 
by a number of students of Oriental farming. The second has been 
frequently noted, but rarely analyzed. The first has been given 
practically no attention. This neglect is particularly unfortunate, 
since the hydraulic patterns of organization and operation have 
decisively affected the managerial role of the hydraulic state. 

Economists generally consider the division of labor and coopera- 
tion key prerequisites of modern industry, but they find them almost 
completely lacking in farming. Their claim reflects the conditions 
of Western rainfall agriculture. For this type of agriculture it is 
indeed by and large correct. 

However, the economists do not as a rule so limit themselves. 
Speaking of agriculture without any geographical or institutional 
qualification, they give the impression that their thesis, being uni- 
versally valid, applies to hydraulic as well as to hydroagriculture and 
rainfall farming. Comparative examination of the facts quickly dis- 
closes the fallacy of this contention. 

a. For early formulations of this view see Smith, 1937: 6; Mill, 1909: 131, 144; Marx, 
DK, I: 300, 322 ff. Modern economists have perpetuated and even sharpened them. 
Writes Seligman (1914: 350): "In the immense domain of agricultural production the 
possibility of combination is almost entirely eliminated." And Marshall (1946: 290): 
"In agriculture there is not much division of labour, and there is no production on 
a very large scale." 


CHAPTER 2, A 23 


1. Preparatory and Protective Operations 
Separated from Farming Proper 

What is true for modern industry — that production proper depends 
on a variety of preparatory and protective operations 6 — has been true 
for hydraulic agriculture since its beginnings. The peculiarity of 
the preparatory and protective hydraulic operations is an essential 
aspect of the peculiarity of hydraulic agriculture. 

a. Large-scale Preparatory Operations [Purpose: Irrigation) 

The combined agricultural activities of an irrigation farmer are 
comparable to the combined agricultural activities of a rainfall 
farmer. But the operations of the former include types of labor 
(on-the-spot ditching, damming, and watering) that are absent in the 
operations of the latter. The magnitude of this special type of labor 
can be judged from the fact that in a Chinese village a peasant may 
spend from 20 to over 50 per cent of his work time irrigating, and 
that in many Indian villages irrigation is the most time-consuming 
single item in the farmer's budget. 1 

Hydroagriculture (small-scale irrigation farming) involves a high 
intensity of cultivation on irrigated fields — and often also on non- 
irrigated fields. 2 But it does not involve a division of labor on a 
communal, territorial, or national level. Such a work pattern occurs 
only when large quantities of water have to be manipulated. Where- 
ever, in pre-industrial civilizations, man gathered, stored, and con- 
ducted water on a large scale, we find the conspicuous division be- 
tween preparatory (feeding) and ultimate labor characteristic of all 
hydraulic agriculture. 

b. Large-scale Protective Operations (Purpose: Flood Control) 

But the fight against the disastrous consequences of too little water 
may involve a fight against the disastrous consequences of too much 
water. The potentially most rewarding areas of hydraulic farming 

b. For the concept of "previous or preparatory labor" see Mill 1909: 29, 31. The 
general principle was already indicated by Smith (1937), who, when discussing the di- 
vision of operations in industry, pointed to the "growers of the flax and the wool" 
and the miners as providers of raw material (5 ff., 11), to the spinners and weavers as 
engaged in special processing operations (6), and to the makers of tools as combining 
elements of both procedures (11). Mill (1909: 36 ff.) also includes, in the category of 
previous labor, activities aimed at protecting industrial production proper. 


are arid and semi-arid plains and humid regions suitable for aquatic 
crops, such as rice, that are sufficiently low-lying to permit watering 
from nearby rivers. These rivers usually have their sources in 
remote mountains, and they rise substantially as the summer sun 
melts part of the snow accumulated there. 

Upstream developments of this kind cause annual inundations in 
Egypt, Mesopotamia, Turkestan, India, China, and in the Andean 
and Mexican zones of America. In semi-arid areas on-the-spot rains 
create additional dangers when they are overconcentrated (con- 
vectional) or irregular. This condition prevails in North China, 
northern Mesopotamia (Assyria), and the Mexican lake region. Thus 
a hydraulic community that resorts to preparatory labor to safe- 
guard the productive use of water may also have to resort to pro- 
tective labor to safeguard its crops from periodic and excessive 

When, in protohistorical times, the Chinese began to cultivate the 
great plains of North China, they quickly recognized that the centers 
of greatest potential fertility were also the centers of greatest po- 
tential destruction. To quote John Lossing Buck: "Geologically 
speaking, man has settled these plains thousands of years before they 
were ready for occupation. . . ." 3 The Chinese built huge em- 
bankments which, although unable to remove entirely the risk 
inhering in the ambivalent situation, matched and even surpassed 
in magnitude the area's preparatory (feeding) works.* 

In India enormous problems of flood control are posed by the 
Indus River 8 and, in a particularly one-sided way, by the Ganges 
and Brahmaputra Rivers, which in Bengal create optimal conditions 
for the cultivation of rice and maximal dangers from floods. By 
1900 Bengal boasted ninety-seven miles of larger irrigation canals 
and 1,298 miles of embankments. 6 

In ancient Mesopotamia even watchful rulers could not com- 
pletely prevent the inundations from damaging the densely settled 
plains. 7 In Turkestan excessive floods periodically threatened the 
Zarafshan River Valley. 8 In Upper Egypt the Nile, in very high flood, 
rises one meter above the level of the settled countryside, in Middle 
Egypt two meters, and in the Delta area up to three and a half 
meters. 9 The inhabitants of the lake area of Mexico could benefit 
from its fertility only if they accepted the periodic overflow of its 
short, irregular, narrow streams, 10 which they sought to control 
through a variety of protective works. Thus in virtually all major 
hydraulic civilizations, preparatory (feeding) works for the purpose 
of irrigation are supplemented by and interlocked with protective 
works for the purpose of flood control. 

chapter 2, a 25 

2. Cooperation 

A study of the hydraulic patterns of China (especially North China), 
India, Turkestan, Mesopotamia (especially Assyria), Egypt, or Meso- 
America (especially the Mexican lake region) must therefore con- 
sider both forms of agrohydraulic activities. Only by proceeding in 
such a way can we hope to determine realistically the dimension 
and character of their organizational key device: cooperation. 

a. Dimension 

When a hydraulic society covers only a single locality, all adult 
males may be assigned to one or a few communal work teams. Vary- 
ing needs and circumstances modify the size of the mobilized labor 
force. In hydraulic countries having several independent sources of 
water supply, the task of controlling the moisture is performed by a 
number of separated work teams. 

Among the Hill Suk of East Africa, "every male must assist 
in making the ditches." 11 In almost all Pueblos "irrigation or clean- 
ing a spring is work for all." 12 Among the Chagga, the maintenance 
of a relatively elaborate irrigation system is assured by "the participa- 
tion of the entire people." 13 In Bali the peasants are obliged to 
render labor service for the hydraulic regional unit, the subak, to 
which they belong. 14 The masters of the Sumerian temple economy 
expected every adult male within their jurisdiction "to participate 
in the digging and cleaning of the canals." 15 Most inscriptions of 
Pharaonic Egypt take this work pattern for granted. Only occasionally 
does a text specify the character of the universally demanded ac- 
tivities, among which lifting and digging are outstanding. 16 

In imperial China every commoner family was expected on de- 
mand to provide labor for hydraulic and other public services. The 
political and legal writings of India indicate a similar claim on 
corviable labor. 17 The laws of Inca Peru obliged all able-bodied men 
to render corvee service. 18 In ancient Mexico both commoner and 
upper-class adolescents were instructed in the techniques of digging 
and damming. 19 At times the masters of this hydraulic area levied 
the manpower of several territorial states for their gigantic hydraulic 
enterprises. 20 

In 19th-century Egypt "the whole corviable population" worked 
in four huge shifts on Mehmed Ali's hydraulic installations. Each 
group labored on the canals for forty-five days until, after 180 
days, the job was completed. 21 From 1881 on, at a time of decay and 
disintegration, "the whole of the corvee fell on the poorest classes," 22 
the smaller number being compensated for by an increase in the 


labor-time to ninety days. In some regions the conscripts were kept 
busy "for 180 days." 


b. Integration 

Orderly cooperation involves planned integration. Such integra- 
tion is especially necessary when the objectives are elaborate and the 
cooperating teams large. 

Above the tribal level, hydraulic activities are usually compre- 
hensive. Most writers who mention the cooperative aspect of hy- 
draulic agriculture think in the main of digging, dredging, and 
damming; and the organizational tasks involved in these labors is 
certainly considerable. But the planners of a major hydraulic enter- 
prise are confronted with problems of a much more complex kind. 
How many persons are needed? And where can such persons be 
found? On the basis of previously made registers, the planners must 
determine the quota and criteria of selection. Notification follows 
selection, and mobilization notification. The assembled groups fre- 
quently proceed in quasimilitary columns. Having reached their 
destination, the buck privates of the hydraulic army must be dis- 
tributed in proper numbers and according to whatever division of 
operations (spading, carrying of mud, etc.) is customary. If raw 
materials such as straw, fagots, lumber, or stone have to be procured, 
auxiliary operations are organized; and if the work teams — in toto 
or in part — must be provided with food and drink, still other ways 
of appropriation, transport, and distribution have to be developed. 
Even in its simplest form, agrohydraulic operations necessitate sub- 
stantial integrative action. In their more elaborate variations, they 
involve extensive and complex organizational planning. 

c. Leadership 

All teamwork requires team leaders; and the work of large inte- 
grated teams requires on-the-spot leaders and disciplinarians as well 
as over-all organizers and planners. The great enterprises of hydraulic 
agriculture involve both types of direction. The foreman usually 
performs no menial work at all; and except for a few engineering 
specialists the sergeants and officers of the labor force are essentially 

To be sure, the physical element — including threats of punish- 
ment and actual coercion — is never absent. But here, if anywhere, re- 
corded experience and calculated foresight are crucial. It is the cir- 
cumspection, resourcefulness, and integrative skill of the supreme 

CHAPTER 2, B 27 

leader and his aides which play the decisive role in initiating, accom- 
plishing, and perpetuating the major works of hydraulic economy. 

d. Hydraulic Leadership — Political Leadership 


The effective management of these works involves an organizational 
web which covers either the whole, or at least the dynamic core, of 
the country's population. In consequence, those who control this net- 
work are uniquely prepared to wield supreme political power. 

From the standpoint of the historical effect, it makes no difference 
whether the heads of a hydraulic government were originally peace 
chiefs, war leaders, priests, priest-chiefs, or hydraulic officials sans 
phrase. Among the Chagga, the hydraulic corvee is called into action 
by the same horn that traditionally rallied the tribesmen for war. 24 
Among the Pueblo Indians the war chiefs (or priests), although sub- 
ordinated to the cacique (the supreme chief), direct and supervise the 
communal activities. 25 The early hydraulic city states of Mesopotamia 
seem to have been for the most part ruled by priest-kings. In China 
the legendary trail blazer of governmental water control, the Great 
Yii, is said to have risen from the rank of a supreme hydraulic func- 
tionary to that of king, becoming, according to protohistorical rec- 
ords, the founder of the first hereditary dynasty, Hsia. 

No matter whether traditionally nonhydraulic leaders initiated or 
seized the incipient hydraulic "apparatus," or whether the masters of 
this apparatus became the motive force behind all important public 
functions, there can be no doubt that in all these cases the resulting 
regime was decisively shaped by the leadership and social control re- 
quired by hydraulic agriculture. 


With regard to operational form, hydraulic agriculture exhibits 
important similarities to heavy industry. Both types of economic ac- 
tivities are preparatory to the ultimate processes of production. Both 

c. Riistow, who in general accepts Kern's view concerning the correlation between 
large-scale and government-directed water control and the centralized and despotic char- 
acter of the state in ancient Egypt and Mesopotamia, assumes that in these areas 
nomadic conquerors developed the hydraulic works after establishing conquest em- 
pires (Riistow, OG, I: 306). 

Patterns of leadership and discipline traditional to conquering groups could be, and 
probably were, invoked in establishing certain hydraulic governments; but Pueblo, 
Chagga, and Hawaiian society show that such formative patterns could also be en- 
dogenous. In any case, the ethnographic and historical facts point to a multiple rather 
than a single origin for hydraulic societies. 


provide the workers with essential material for these ultimate proc- 
esses. And both tend to be comprehensive, "heavy." For these rea- 
sons the large enterprises of hydraulic agriculture may be designated 
as "heavy water works." 

But the dissimilarities are as illuminating as the similarities. The 
heavy water works of hydraulic agriculture and the heavy industry 
of modern economy are distinguished by a number of basic differ- 
ences, which, properly denned, may aid us in more clearly recogniz- 
ing the peculiarities of hydraulic society. 

Heavy water works feed the ultimate agrarian producer one cru- 
cial auxiliary material: water; heavy industry provides auxiliary and 
raw materials of various kinds, including tools for finishing and heavy 
industry. Heavy water works fulfill important protective functions 
for the country at large; the protective installations (buildings, etc.) 
of industry do not. Heavy water works cover at their inception a rel- 
atively large area; and with the development of the hydraulic order 
they are usually spread still further. The operations of heavy indus- 
try are spatially much more restricted. At first, and for a number of 
preliminary processes, they may depend on small and dispersed shops; 
with the growth of the industrial order they tend to merge into one, 
or a few, major establishments. 

The character of the labor force varies with these spatial and op- 
erational differences. Heavy water works are best served by a widely 
distributed personnel, whereas heavy industry requires the workers 
to reside near the locally restricted "big" enterprises which employ 
them. The hydraulic demand is satisfied by adult peasant males, who 
continue to reside in their respective villages; whereas the industrial 
demand is satisfied by a geographically concentrated labor force. 

The bulk of the hydraulic workers are expected to remain peas- 
ants, and in most cases they are mobilized for a relatively short 
period only — at best for a few days, at worst for any time that will 
not destroy their agricultural usefulness. Thus division of agrohy- 
draulic labor is not accompanied by a corresponding division of 

The contrast to the labor policy of heavy industry is manifest. Dif- 
ferent from heavy water works, which may be created and maintained 
during a fraction of the year, heavy industry operates most effectively 
when it operates continuously. The industrial employers prefer to 
occupy their personnel throughout the year; and with the growth of 
the industrial system full-time labor became the rule. Thus division 
of industrial labor moves toward a more or less complete division of 

The two sectors are also differently administered. In the main, 

CHAPTER 2, C 29 

modern heavy industry is directed by private owners or managers. 
The heavy water works of hydraulic agriculture are directed es- 
sentially by the government. The government also engages in certain 
other large enterprises, which, in varying combinations, supplement 
the agrohydraulic economy proper. 


Among the intellectual functions fulfilled by the leaders of agro- 
hydraulic activities, some are only indirectly connected with the 
organization of men and material; but the relation is highly signifi- 
cant nevertheless. Time keeping and calendar making are essential 
for the success of all hydraulic economies; and under special con- 
ditions special operations of measuring and calculating may be 
urgently needed. 1 The way in which these tasks are executed affect 
both the political and the cultural development of hydraulic society. 

To be sure, man is deeply concerned about the swing of the seasons 
under all forms of extractive economy and throughout the agrarian 
world. But in most cases he is content to determine in a general way 
when spring or summer begin, when cold will set in, when rain or 
snow will fall. In hydraulic civilizations such general knowledge is 
insufficient. In areas of full aridity it is crucial to be prepared for 
the rise of the rivers whose overflow, properly handled, brings 
fertility and life and whose unchecked waters leave death and 
devastation in their wake. The dikes have to be repaired in the 
proper season so that they will hold in times of inundation; and 
the canals have to be cleaned so that the moisture will be satis- 
factorily distributed. In semi-arid areas receiving a limited or uneven 
rainfall an accurate calendar is similarly important. Only when the 
embankments, canals, and reservoirs are ready and in good condition 
can the scanty precipitation be fully utilized. 

The need for reallocating the periodically flooded fields and 
determining the dimension and bulk of hydraulic and other struc- 
tures provide continual stimulation for developments in geometry 
and arithmetic. Herodotus ascribes the beginnings of geometry in 
Egypt to the need for annually remeasuring the inundated land. 2 

No matter whether the earliest scientific steps in this direction were 
made in the Nile Valley or in Mesopotamia, the basic correlation is 
eminently plausible. Obviously the pioneers and masters of hydraulic 
civilization were singularly well equipped to lay the foundations for 
two major and interrelated sciences: astronomy and mathematics. 

As a rule, the operations of time keeping and scientific measuring 


and counting were performed by official dignitaries or by priestly 
(or secular) specialists attached to the hydraulic regime. Wrapped in 
a cloak of magic and astrology and hedged with profound secrecy, 
these mathematical and astronomical operations became the means 
both for improving hydraulic production and bulwarking the 
superior power of the hydraulic leaders. 


The masters of the hydraulic state did not confine their activities to 
matters immediately connected with agriculture. The methods of 
cooperation which were so effective in the sphere of crop-raising 
were easily applied to a variety of other large tasks. 

Certain types of works are likely to precede others. Generally 
speaking, the irrigation canal is older than the navigation canal; and 
hydraulic digging and damming occurred prior to the building of 
highways. But often derivative steps were taken before the original 
activities had progressed far, and different regional conditions favored 
different evolutionary sequences. Thus the divergencies of inter- 
action and growth are great. They include many constructional 
activities ab©ve and beyond the sphere of hydraulic agriculture.* 


a. Aqueducts and Reservoirs Providing Drinking Water 

A commonwealth able to transfer water for purposes of irriga- 
tion readily applies its hydraulic know-how to the providing of 
drinking water. The need for such action was slight in the greater 
part of Medieval Europe, where the annual precipitation furnished 
sufficient ground water for the wells on which most towns depended 
for their water supply. 1 

Even in the hydraulic world, drinking water is not necessarily 
an issue. Wherever rivers, streams, or springs carry enough moisture 

a. Anyone interested in studying the technical and organizational details of a major 
hydraulic order may consult Willcocks' admirable description of irrigation and flood 
control in 19th-century Egypt (Willcocks, 1889: passim). A comprehensive survey of the 
hydraulic conditions in India at the close of the 19th century has been made by the 
Indian Irrigation Commission (RRCAI). In my study of Chinese economics and society 
I have systematically analyzed the ecological foundations and the various aspects of 
China's traditional hydraulic order (Wittfogel, 1931: 61-93, 188-300, and 410-56). Today 
we also have an archaeological account of the growth of hydraulic and other construc- 
tions over time and for a limited, but evidently, representative area: the Virii Valley 
in Peru (see Willey, 1953: 344-89)- 

CHAPTER 2, D £1 

to satisfy the drinking needs of the population throughout the year, 
no major problem arises. The inhabitants of the Nile and Ganges 
Valleys and of many similar areas did not have to construct elaborate 
aqueducts for this purpose. 

The irregular flow of rivers or streams or the relatively easy access 
to fresh and clear mountain water has stimulated in many hydraulic 
landscapes the construction of comprehensive installations for the 
storage and distribution of drinking water. In America great aque- 
ducts were built by the hydraulic civilizations of the Andean zone 
and Meso-America. 2 The many reservoirs (tanks) of Southern India 
frequently serve several uses; but near the large residential centers 
the providing of drinking water is usually paramount. In certain 
areas of the Near East, such as Syria and Assyria, brilliantly designed 
aqueducts have satisfied the water needs of many famous cities, 
Tyre, 3 Antioch, 4 and Nineveh 5 among them. In the Western world 
of rainfall agriculture, aqueducts were built primarily by such 
Mediterranean peoples as the Greeks and the Romans, who since 
the dawn of history maintained contact with — and learned from — 
the technically advanced countries of Western Asia and North Africa. 
No doubt the Greeks and Romans would have been able to solve 
their drinking-water problem without inspiration from the outside; 
but the form of their answer strongly suggests the influence of 
Oriental engineering. 6 

b. Navigation Canals 

Among the great agrarian conformations Of history, only hydraulic 
society has constructed navigation canals of any major size. The 
seafaring Greeks, making the Mediterranean their highway, avoided 
an issue which the ancient city states were poorly equipped to handle. 
The not-too-numerous Roman canals were apparently all dug at a 
time when the growing Orientalization of the governmental ap- 
paratus stimulated, among other things, a growing interest in all 
kinds of public works. 7 

The rainfall farmers of Medieval Europe, like their counterparts 
elsewhere, shunned rather than sought the marshy river lowlands. 
And their feudal masters paid little attention to the condition of the 
watercourses, for which they had no use. Still less did they feel 
obliged to construct additional and artificial rivers — canals. Few if 
any important canals were built during the Middle Ages, 8 and 
medieval trade and transport were seriously handicapped by the 
state of the navigable rivers. 9 

It was in connection with the rise of a governmentally encouraged 


commercial and industrial capitalism that the West began to build 
canals on a conspicuous scale. The "pioneer of the canals of modern 
Europe," the French Canal du Midi, was completed only in the 
second half of the 17th century, in 1681, 10 that is, little more than a 
century before the end of the absolutist regime. And in the classical 
country of inland navigation, England, 11 "little . . . was done in 
making canals . . . until the middle of the eighteenth century" 12 — 
that is, until a time well after the close of England's absolutist period 
and immediately prior to the beginning of the machine age. 

As stated above, the members of a hydraulic commonwealth felt 
quite differently about the management of natural and artificial 
watercourses. They approached the fertility-bearing rivers as closely 
as possible, and in doing so they had to find ways of draining the 
lowland marshes and strengthening and reshaping the river banks. 
Naturally the question of inland navigation did not arise everywhere. 
Existing rivers and streams might be suitable for irrigation, but not 
for shipping (Pueblos, Chagga, Highland Peru); or the ocean might 
prove an ideal means of transportation (Hawaii, Coastal Peru). In 
certain localities inland navigation was satisfactorily served by man- 
managed rivers (Egypt, India) and lakes (Mexico) plus whatever ir- 
rigation canals were large enough to accommodate boats (Meso- 

But when supplementary watercourses were not only possible but 
desirable, the organizers of agrohydraulic works had little difficulty 
in utilizing their cooperative "apparatus" to make them available. 
The new canals might be only minor additions to the existing 
watercourses. The ancient Egyptians constructed canals in order to 
circumnavigate impassable cataracts, and they temporarily connected 
the Nile and the Red Sea; 13 but these enterprises had little effect 
on the over-all pattern of the country's hydraulic economy. In other 
instances, navigation canals assumed great importance. They satisfied 
the needs of the masters of the hydraulic state: the transfer of parts 
of the agrarian surplus to the administrative centers and the trans- 
port of messengers and troops. 

In Thailand (Siam) the different hydraulic tasks overlapped. In 
addition to the various types of productive and protective hydraulic 
installations, the government constructed in the centers of rice 
production and state power a number of canals, which essentially 
served as "waterways," that is, as a means for transporting the rice 
surplus to the capital. 14 

The corresponding development in China is particularly well 
documented. In the large plains of North China the beginnings of 
navigation canals go back to the days of the territorial states — that 

CHAPTER 2, D 33 

is, to the period prior to 221 B.C., when the various regional govern- 
ments were still administered by officials who were given office lands 
in payment for their services. The difference between the state- 
centered system of land grants as it prevailed in early China and 
the knighthood feudalism of Medieval Europe is spectacularly dem- 
onstrated by the almost complete absence of public works in feudal 
Europe and the enormous development of such works — hydraulic 
and otherwise — in the territorial states of China. 6 

The geographical and administrative unification of China which 
vastly increased the political need for navigation canals also in- 
creased the state's organizational power to build them. The first 
centuries of the empire saw a great advance not only in the con- 
struction of irrigation canals, 15 reservoirs, and protective river dikes 
but also in the digging of long canals for administrative and fiscal 
purposes. 18 

When, after several centuries of political fragmentation, the Sui 
rulers at the end of the 6th century again unified "all-under-heaven," 
they bulwarked the new political structure by creating out of earlier 
and substantial beginnings the gigantic Imperial Canal, significantly 
known in China as Yiin Ho, "the Transport Canal." This canal ex- 
tends today for about 800 miles, its length equaling the distance 
from the American-Canadian Great Lakes to the Gulf of Mexico or 

b. Previously I viewed Chou China as a feudal society exhibiting Oriental features, 
which appeared early and became increasingly conspicuous until, at the close of the 
period, they prevailed completely (Wittfogel, 1931: 278 ft.; ibid., 1935: 40 ft.). The idea 
of a society that crosses the institutional divide is entirely compatible with the findings 
of the present inquiry (see below, Chap. 6); and by interpreting Chou society in this 
way, I would not have had to change a long-held position. But intensified comparative 
studies compel me to change. The arid and semi-arid settings of North China (17 
inches annual rainfall in the old Chou domain and 24 inches in the domain of the 
pre-Chou dynasty, Shang) suggest hydraulic agriculture for the ancient core areas. 
The lay of the land, the summer floods, and the periodic silting-up of the rivers neces- 
sitated comprehensive measures of flood control especially in the heartland of Shang 
power. A realistic interpretation of legends and protohistorical sources (cf. Wittfogel 
and Goldfrank, 1943: passim) points to the rise of a hydraulic way of life long before 
the Shang dynasty, whose artifacts (bronzes) and inscriptions reflect a highly developed 
agrarian civilization with refined techniques of record keeping, calculations, and astron- 
omy. The recognizable institutions of early Chou are those of a hydraulic society, 
which gradually intensified its managerial and bureaucratic "density" (for this con- 
cept see below, Chap. 6). The Chou sovereigns behaved toward the territorial rulers 
not as the first among equals but as supreme masters responsible only to Heaven. It 
was not their fault that their despotic claims, which possibly imitated Shang precedents, 
were realized imperfectly and with decreasing effect. In contrast, the rulers of the ter- 
ritorial states were strong enough to proceed absolutistically within their respective 
realms. The lands that they assigned were given not in a contractual way and to in- 
dependently organized (corporated) knights and barons, but to office holders and persons 
permitted to enjoy sinecures. They were not fiefs but office lands (see below, Chaps. 6-8). 


— in European terms — the distance from Berlin to Bordeaux or from 
Hamburg to Rome. For labor on part o£ this gigantic water work 
the Sui government mobilized in the regions north of the Yellow 
River alone "more than a million of men and women," " that is, 
almost one-half of the total population which England is said to have 
had from the 14th to the 16th century. 18 

The gigantic effort involved in banking the rivers and building 
the canals of China is indicated by the American agronomist, F. H. 
King, who conservatively estimates the combined lengths of the 
man-managed watercourses of China, Korea, and Japan at some 
200,000 miles. "Forty canals across the United States from east to 
west and sixty from north to south would not equal in number of 
miles those in these three countries today. Indeed, it is probable that 
this estimate is not too large for China alone." 


2. Large Non/iydraulic Constructions 

a. Huge Defense Structures 

The need for comprehensive works of defense arises almost as soon 
as hydraulic agriculture is practiced. Contrary to the rainfall farmer, 
who may shift his fields with relative ease, the irrigation farmer 
finds himself depending on an unmovable, if highly rewarding, 
source of fertility. In the early days of hydraulic cultivation reliance 
on a fixed system of water supply must in many cases have driven 
the agrarian community to build strong defenses around its homes 
and fields. 

For this purpose hydraulic agriculture proved suggestive in two 
ways: it taught man how to handle all kinds of building materials, 
earth, stone, timber, etc., and it trained him to manipulate these 
materials in an organized way. The builders of canals and dams 
easily became the builders of trenches, towers, palisades, and ex- 
tended defense walls. 

In this, as in all corresponding cases, the character and magnitude 
of the operations were determined by internal and external cir- 
cumstances. Surrounded by aggressive neighbors, the Pueblo Indians 
ingeniously utilized whatever building material was at hand to 
protect their settlements, which rarely comprised more than a few 
hundred inhabitants." The fortress-like quality of their villages is 
manifest to the present-day anthropologist; it struck the Spanish 

c. Castaneda, 1896: 512. Bandelier upholds Castaneda's figures against divergent 
statements made in other early Spanish sources (Bandelier, FR, I: 120 ff. and nn.; cf. 
ibid., DH: 312, 46 ft., 171-3). 

CHAPTER 2, D 35 

conquistador es, who were forced at times to besiege a single settle- 
ment for days and weeks before they could take it. d Rigid coopera- 
tion assured security of residence, just as it assured success in farm- 
ing. An early observer stresses this aspect of Pueblo life: "They all 
work together to build the villages." * 

d. Castaneda, who was the official chronicler of the first Spanish expedition, notes 
(1896: 494) that the defense towers of a large Zuni settlement were equipped with 
"embrassures and loopholes ... for defending the roofs of the different stories." He 
adds, "The roofs have to be reached first, and these upper houses are the means 
of defending them." The experiences of the second expedition confirmed and supple- 
mented the initial observations. Gallegos concludes his remarks concerning Pueblo 
building by referring to the movable wooden ladders "by means of which they climb 
to their quarters." At night "they lift them up since they wage war with one another" 
(Gallegos, 1927: 265). Obregon also stresses the military value of the ladders; in 
addition, he explains how the edifices themselves served to protect the community: 
"These houses have walls and loopholes from which they defend themselves and 
attack their enemies in their battles" (Obregon, 1928: 293). 

One of Coronado's lieutenants, approaching certain Tigua settlements, "found the 
villages closed by palisades." The Pueblos, whose inhabitants had been subjected to 
various forms of extortion and insult "were all ready for fighting. Nothing could be 
done, because they would not come down onto the plain and the villages are so strong 
that the Spaniards could not dislodge them." Attacking a hostile village, the Spanish 
soldiers reached the upper story by surprise tactics. They remained in this dangerous 
position for a whole day, unable to prevail until the Mexican Indians, who accom- 
panied them, approached the Pueblo from below, digging their way in and smoking out 
the defenders (Castaneda, 1896: 496. For a discussion of Castaneda's report see Bandelier, 
DH: 38 fL). 

Besieging a large Tigua settlement, Coronado's men had an opportunity to test 
thoroughly the defense potential of a Pueblo which was not taken by surprise: "As the 
enemy had had several days to provide themselves with stores, they threw down 
such quantities of rocks upon our men that many of them were laid down, and they 
wounded nearly a hundred with arrows." The siege lasted for seven weeks. During 
this time, the Spaniards made several assaults; but they were unable to take the 
Pueblo. The villagers eventually abandoned their fortress-like bulwark, not because 
the aggressors had penetrated their defenses, but because of lack of water (Castaneda, 
1896: 498 fL; cf. RDS: 576). Bandelier supplements Castaneda's report of this significant 
event by an account given by Mota Padilla, an 18th-century author, who claims to have 
had access to the original writings of still another member of Coronado's staff 
(Bandelier, DH: 323). Mota Padilla's version contains a number of details which 
reveal the techniques of attack as well as the strength and ingenuity of the defense. 
Some of the Spaniards "reached the top of the wall, but there they found that the 
natives had removed the roofs of many (upper) rooms, so that there was no communica- 
tion between them, and as there were little towers at short distances from each other, 
from which missiles were showered upon the assailants on the top, the Spaniards had 
more than sixty of their number hurt, three of whom died of their wounds" (ibid., 48). 

e. Castaneda (1896: 520) qualifies this general statement by saying that the women 
were "engaged in making the [adobe] mixture and the walls, while the men bring 
the wood and put it in place." Modern reports assign the above duties to the men 
and credit them in addition with erecting the walls, the construction labors of the 
women being confined to plastering (White, 1932: 33; cf. Parsons, 1932: 212). The 


The Chagga were equally effective in the transfer of their hydraulic 
work patterns to military constructions. Their great chieftain, 
Horombo (/?. 1830), used "thousands of people" to build great 
fortifications, which in part still stand today. 20 "The walls of these 
fortifications are some six feet high, and in length 305 yards on the 
south side, 443 yards on the north, 277 yards on the east side, and 
137 yards on the west side." 21 Tunnels, extended trenches, and dug- 
outs added to the defense of the walled settlements, which appeared 
early in the history of the Chagga. 22 "Deep dugouts excavated under 
the huts and often leading into underground passages with outlets 
at some distance, were used for refuge. Almost every country was 
secured with great war trenches, which are everywhere to be seen 
at the present day and are often still of great depth." 23 

These instances show what even primitive hydraulic societies could 
achieve in the field of defense construction, when they strained their 
cooperative resources to the full. Higher hydraulic societies em- 
ployed and varied the basic principle in accordance with technical 
and institutional circumstances. 

In pre-Columbian Mexico the absence of suitable labor animals 
placed a limitation on transport, and while this restricted siege craft, 
it did not preclude the struggle for or the defense of the cities. In 
emergencies many government-built hydraulic works in the main 
lake area fulfilled military functions, just as the monster palaces and 
temples served as bastions against an invading enemy. 2 * Recent re- 
search draws attention to various types of Mexican forts and defense 
walls. 25 Because of their size and importance, they may safely be 
adjudged as state-directed enterprises. The colossal fortresses and 
walls of pre-Spanish Peru, which astonished early and recent ob- 
servers, 26 are known to have been built at the order of the govern- 
ment and by "incredibly" large teams of corvee laborers. 27 

Many texts and pictorial representations have portrayed the walls, 
gates, and towers of ancient Egypt, Sumer, Babylonia, Assyria, and 
Syria. The Arthashdstra indicates the systematic manner in which the 
rulers of the first great Indian empire treated problems of fortifica- 
tion and defense. 28 At the dawn of Chinese history new capitals were 
created at the ruler's command, and during the last centuries of the 
Chou period the territorial states used their corviable manpower to 
wall entire frontier regions, not only against the tribal barbarians 
but also against each other. In the 3d century b.c. the unifier of 

divergence between the early and recent descriptions may reflect an actual institutional 
change or merely a difference in the accuracy of observation. While interesting to the 
anthropologist, this discrepancy does not affect our basic conclusions regarding the 
communal character of large-scale building in the American Pueblos. 

CHAPTER 2, D 37 

China, Ch'in Shih Huang-ti, linked together and elaborated older 
territorial structures to form the longest unbroken defense installa- 
tion ever made by man. 29 The periodic reconstruction of the Chinese 
Great Wall expresses the continued effectiveness of hydraulic econ- 
omy and government-directed mass labor. 

b. Roads 

The existence of government-made highways is suggested for the 
Babylonian period; 30 it is documented for Assyria. 31 And the rela- 
tionship between these early constructions and the roads of Persia, 
the Hellenistic states, and Rome seems "beyond doubt." ; The great 
Persian "royal road" deeply impressed the contemporary Greeks; 32 
it served as a model for the Hellenistic rulers, 33 whose efforts in 
turn inspired the official road builders of the Roman empire. 34 Ac- 
cording to Mez, the Arabs inherited "the type of 'governmental road,' 
like its name, from the Persian 'Royal Road.' " 35 Beyond this, how- 
ever, they showed little interest in maintaining good roads, probably 
because they continued to rely in the main on camel caravans for 
purposes of transport. The later Muslim regimes of the Near East 
used highways, but they never restored them to the state of technical 
perfection which characterized the pre-Arab period. 38 

Roads were a serious concern of India's vigorous Maurya kings. 37 
A "royal road" of 10,000 stadia, which is said to have led from the 
capital to the northwestern border, had a system of marking dis- 
tances which, in a modified form, was again employed by the Mogul 
emperors. 38 In Southern India, where Hindu civilization was per- 
petuated for centuries after the north had been conquered, govern- 
ment-made roads are mentioned in the inscriptions; and "some of 
them are called king's highways." 39 The Muslim rulers of India 
continued the Indian rather than the West Asian pattern in their 
effort to maintain a network of state roads. 40 Sher Shah (d. 1545) 
built four great roads, one of which ran from Bengal to Agra, Delhi, 
and Lahore. 41 Akbar is said to have been inspired by Sher Shah 
when he built a new "king's highway," called the Long Walk, which 
for four hundred miles was "shaded by great trees on both sides." 42 

In China, a gigantic network of highways was constructed im- 
mediately after the establishment of the empire in 221 B.C. But in 
this case, as in the cases of the irrigation and navigation canals or 

/. Meissner, BA, I: 341. The term "royal road" was used in an Assyrian inscription 
(Olmstead, 1923: 334). The operational pattern of the Roman state post, the cursus 
publicus, can be traced back through the Hellenistic period to Persia and perhaps even 
to Babylonia (Wilcken, 1912: 372 and n. 2). 



the long defense walls, the imperial engineers systematized and 
elaborated only what their territorial predecessors had initiated. Long 
before the 3d century b.c. an efficient territorial state was expected 
to have well kept overland highways, supervised by central and 
local officials, lined with trees, and provided with stations and guest 
houses. 43 Under the empire, great state roads connected all the im- 
portant centers of the northern core area with the capital. Accord- 
ing to the official History of the Han Dynasty, the First Emperor 

built the Imperial Road throughout the empire. To the east it 
stretched to Yen and Ch'i and to the south it reached Wu and 
Gh'u. The banks and the shore of the Chiang [the Yangtze 
River] and the lakes and the littoral along the sea coast were 
all made accessible. The highway was fifty paces wide. A space 
three chang [approximately twenty-two feet] wide in the center 
was set apart by trees. The two sides were firmly built, and 
metal bars were used to reinforce them. Green pine trees were 
planted along it. He constructed the Imperial Highway with 
such a degree of elegance that later generations were even unable 
to find a crooked path upon which to place their feet, 44 

In the subsequent dynasties the building and maintenance of the 
great trunk roads and their many regional branches remained a 
standard task of China's central and local administration. 

The rugged terrain of Meso-America and the absence of fully 
coordinated empires seems to have discouraged the construction of 
highways during the pre-Columbian period, at least on the high 
plateau. But the Andean area was the scene of extraordinary road 
building. The Spanish conquerors described in detail the fine high- 
ways which crossed both the coastal plain and the highlands and 
which formed connecting links between them. 45 Commenting on the 
Andean roads, Hernando Pizarro writes he never saw their like in 
similar terrain "within the entire Christian world." 46 In fact the 
only parallel he could think of was the system of highways built by 
the Romans. The similarity is telling. As we shall discuss below, the 
extensive Roman roads were the fruits of a fateful transformation 
that made the Roman Empire a Hellenistically (Orientally) despotic 

The efforts required to build all these great highways have at- 
tracted much less attention than the finished products. But what 
evidence we have indicates that like most other major government 
enterprises, they were mainly executed through the cooperative effort 
of state-levied corvee laborers. Under the Inca empire supervisory 

CHAPTER 2, D 39 

officials marked off the land and informed the local inhabitants "that 
they should make these roads." And this was done with little cost 
to the government. The commandeered men "come with their food 
and tools to make them." " 

The highways of imperial China required an enormous labor 
force for their construction and a very sizable one for their mainte- 
nance. A Han inscription notes that the construction of a certain 
highway in the years a.d. 63-66 occupied 766,800 men. Of this great 
number only 2,690 were convicts.* 7 

c. Palaces, Capital Cities, and Tombs 

A governmental apparatus capable of executing all these hydrau- 
lic and nonhydraulic works could easily be used in building palaces 
and pleasure grounds for the ruler and his court, palace-like govern- 
ment edifices for his aides, and monuments and tombs for the dis- 
tinguished dead. It could be used wherever the equalitarian condi- 
tions of a primitive tribal society yielded to tribal or no-longer 
tribal forms of autocracy. 

The head chief of a Pueblo community had his fields worked for 
him by the villagers. But apparently his dwelling did not differ 
from the houses of other tribesmen, except perhaps that it was 
better and more securely located. The Chagga chieftains had veri- 
table palaces erected for their personal use; and the corvee labor in- 
volved in their construction was substantial. 48 

The colossal palaces of the rulers of ancient Peru were erected by 
the integrated manpower of many laborers. In pre-Columbian 
Mexico, Nezahualcoyotzin, the king of Tezcuco, the second largest 
country in the Aztec Federation, is said to have employed more than 
200,000 workers each day for the building of his magnificent palace 
and park. 49 

Unlimited control over the labor power of their subjects enabled 
the rulers of Sumer, Babylon, and Egypt to build their spectacular 
palaces, gardens, and tombs. The same work pattern prevailed in 
the many smaller states that shaped their government on the Meso- 
potamian or Egyptian model. According to the biblical records, 
King Solomon built his beautiful temple with labor teams that, like 
those of Babylonia, were kept at work for four months of the year. 


g. Cieza, 1943: 95. The regional organization and the repair work on the roads had 
already been noted by a member of the conquering army (Estete, 1938: 246). The 
lack of payment for services rendered in the road corvee is also recorded by Bias 
Valeras, who states that similar conditions prevailed with regard to work on the 
bridges and irrigation canals (Garcilaso, 1945, I: 258). 


The great edifices of Mogul India have been frequently described. 
Less known but equally worthy of mention are the constructions of 
the earlier periods. The third ruler of the Tughluq, Firus Shah (ca. 
1308-88), dug several important irrigation canals, the famous "Old 
Jumna Canal" among them. He built forts, palaces, and palace-cities, 
mosques, and tombs. The palace-fort of Kotla Firus Shah, which rose 
in his new capital of FIrusabad (Delhi), faithfully preserved the 
grand style of pre-Islamic Indian and Eastern architecture. 51 

The Chinese variant of the general agromanagerial building trend 
is revealed in many elaborate works. The First Emperor of China, 
Ch'in Shih Huang-ti, began to build great hydraulic works in the 
early days of his power; and in the course of his reign he completed 
colossal works of the nonhydraulic public and semiprivate types. 
Having destroyed all his territorial rivals, he constructed the previ- 
ously mentioned network of highways which gave his officials, 
messengers, and troops easy access to all regions of his far-flung 
empire. Later he defended himself against the northern pastoralists 
by consolidating the Great Wall. Palaces for his personal use had 
been built in the early days of his reign; but it was only in 213 b.c. 
that work was begun on his superpalace. This monster project, to- 
gether with the construction of his enormous tomb, 52 is said to have 
occupied work teams numbering over 700,000 persons. 53 

Eight hundred years later the second monarch of a reunified 
China, Emperor Yang (604-17) of the Sui Dynasty, mobilized a still 
larger labor force for the execution of similar monster enterprises. 
In addition to the more than one million persons — men and women 
— levied for the making of the Grand Canal, 54 he dispatched huge 
corvee teams to extend the imperial roads 55 and to work on the 
Great Wall. According to the History of the Sui Dynasty, over a 
million persons toiled at the Great Wall.* According to the same 
official source, the construction of the new eastern capital, which 
included a gigantic new imperial palace, involved no less than two 
million people "every month." " 

d. Temples 

The position, fate, and prestige of the secular masters of hydraulic 
society were closely interlinked with that of their divine protectors. 
Without exception, the political rulers were eager to confirm and 
bulwark their own legitimacy and majesty by underlining the great- 
ness of their supernatural supporters. Whether the government was 

h. Over a million in 607; an additional 200,000 persons were employed in 608 
(Sui Shu 3. 10b, 12a). 

CHAPTER 2, D 41 

headed by secular monarchs or priest-kings, the commanding center 
made every effort to provide the supreme gods and their earthly 
functionaries with adequate surroundings for worship and residence. 

Government-directed work teams, which erected gigantic palaces, 
were equally fitted to erect gigantic temples. Ancient inscriptions 
note the many temples built by the Mesopotamian rulers. 57 Usually 
the sovereign speaks as if these achievements resulted solely from his 
personal efforts. But occasional remarks indicate the presence of "the 
people" who toiled "according to the established plan." * Similarly, 
most Pharaonic texts refer to the final achievement s or to the great- 
ness of the directing sovereign; 88 but again a number of texts refer 
to the government-led labor forces, "the people." * 

In the agromanagerial cultures of pre-Columbian America, build- 
ings for religious purposes were particularly conspicuous. Native 
tradition as well as the early Spanish accounts emphasize the tremen- 
dous labor required to construct and maintain the sacred houses 
and pyramids. The Mexicans coordinated their communal energies 
to erect the first temple for the newly established island city, the later 
Aztec capital; 59 and their increasingly powerful descendants mobi- 
lized the manpower of many subjugated countries for the construc- 
tion of increasingly huge temples." 1 The city-like palace of the famous 
King of Tezcuco, Nezahualcoyotzin, contained no less than forty 
temples. 00 The great number of laborers engaged in building this 
palace- and temple-city has already been cited. Like the monster 
work teams of Mexico, those of Tezcuco could draw upon the entire 
corviable population." In another country of the main lake region, 
Cuauhtitlan, the construction of large-scale hydraulic works 61 was 
followed by the building of a great temple. It took thirteen years to 
complete the second task. 62 

In the Andean zone, as in most other areas of the hydraulic world, 
the attachment of the priesthood to the government is beyond doubt. 
The Incas made heavy levies on their empire's material wealth in 

i. Price, 1927: 24; cf. Thureau-Dangin, 1907: 111, and Barton, 1929: 225. Schneider 
(1920: 46) and Deimel (1931: 101 ff.) deplore the scarcity of concrete data concerning 
the Sumerian construction industry. 

;'. Thus in one of the oldest inscriptions of Egypt extant, the Palermo Stone 
(Breasted, 1927, I: 64). 

k. "I have commanded those who work, to do according as thou shalt exact" 
(Breasted, 1927, I: 245). The "people" bring the stone for the Amon Temple; and 
the "people" also do the building. Among the workmen are several types of artisans 
(ibid., II: 294, 293). 

m. Tezozomoc, 1944: 79 (the Temple of Huitzilopochtli) and 157 (the great Cu 
edifice of the same god). 

n. Ixtlilxochitl, OH, II: 173 ff. The Annals of Cuauhtitlan also refer to this construc- 
tion (Chimalpopoca, 1945: 52), without, however, discussing the labor aspect. 


order to beautify their temples and pyramids. 63 They called up what- 
ever manpower was needed to collect the raw material, transport it, 
and do the actual work of construction. 64 


Evidently the masters of hydraulic society, whether they ruled in 
the Near East, India, China, or pre-Conquest America, were great 
builders. The formula is usually invoked for both the aesthetic and 
the technical aspect of the matter; and these two aspects are indeed 
closely interrelated. We shall briefly discuss both of them with regard 
to the following types of hydraulic and nonhydraulic construction 

I. Hydraulic works 

A. Productive installations 

(Canals, aqueducts, reservoirs, sluices, and dikes for the pur- 
pose of irrigation) 

B. Protective installations 

(Drainage canals and dikes for flood control) 

C. Aqueducts providing drinking water 

D. Navigation canals 

II. Nonhydraulic works 

A. Works of defense and communication 

1 . Walls and other structures of defense 

2. Highways 

B. Edifices serving the public and personal needs of the secular 
and religious masters of hydraulic society 

i. Palaces and capital cities 

2. Tombs 

3. Temples 

1. The Aesthetic Aspect 

a. Uneven Conspicuousness 

The majority of persons who have commented on the great builders 
of Asia and ancient America are far more articulate on the non- 
hydraulic than on the hydraulic achievements. Within the hydraulic 
sphere more attention is again given to the aqueducts for drinking 
water and the navigation canals than to the productive and protec- 
tive installations of hydraulic agriculture. In fact, these last are fre- 

CHAPTER 2, E 43 

quently overlooked altogether. Among the nonhydraulic works, the 
"big houses" of power and worship and the tombs of the great are 
much more carefully investigated than are the large installations of 
communication and defense. 

This uneven treatment of the monster constructions of hydraulic 
society is no accident. For functional, aesthetic, and social reasons the 
hydraulic works are usually less impressive than the nonhydraulic 
constructions. And similar reasons encourage uneven treatment also 
within each of the two main categories. 

Functionally speaking, irrigation canals and protective embank- 
ments are widely and monotonously spread over the landscape, 
whereas the palaces, tombs, and temples are spatially concentrated. 
Aesthetically speaking, most of the hydraulic works are undertaken 
primarily for utilitarian purposes, whereas the residences of the rulers 
and priests, the houses of worship, and the tombs of the great are 
meant to be beautiful. Socially speaking, those who organize the 
distribution of manpower and material are the same persons who 
particularly and directly enjoy the benefits of many nonhydraulic 
structures. In consequence they are eager to invest a maximum of 
aesthetic effort in these structures (palaces, temples, and capital 
cities) and a minimum of such effort in all other works. 

Of course, the contrast is not absolute. Some irrigation works, 
dikes, aqueducts, navigation canals, highways, and defense walls do 
achieve considerable functional beauty. And closeness to the centers 
of power may lead the officials in charge to construct embankments, 
aqueducts, highways, bridges, walls, gates, and towers with as much 
care for aesthetic detail as material and labor permit. 

But these secondary tendencies do not alter the two basic facts 
that the majority of all hydraulic and nonhydraulic public works 
are aesthetically less conspicuous than the royal and official palaces, 
temples, and tombs, and that the most important of all hydraulic 
works — the canals and dikes — from the standpoint of art and artistry 
are the least spectacular of all. 

b. The Monumental Style 

Such discrepancies notwithstanding, the palaces, government build- 
ings, temples, and tombs share one feature with the "public" works 
proper: they, too, tend to be large. The architectural style of hydraulic 
society is monumental. 

This style is apparent in the fortress-like settlements of the Pueblo 
Indians. It is conspicuous in the palaces, temple cities, and fortresses 
of ancient Middle and South America. It characterizes the tombs, 


palace-cities, temples, and royal monuments of Pharaonic Egypt and 
ancient Mesopotamia. No one who has ever observed the city gates 
and walls of a Chinese capital, such as Peking, or who has walked 
through the immense palace gates and squares of the Forbidden City 
to enter its equally immense court buildings, ancestral temples, and 
private residences can fail to be awed by their monumental design. 
Pyramids and dome-shaped tombs manifest most consistently the 
monumental style of hydraulic building. They achieve their aesthetic 
effect with a minimum of ideas and a maximum of material. The 
pyramid is little more than a huge pile of symmetrically arranged 

The property-based and increasingly individualistic society of an- 
cient Greece loosened up the massive architecture, which had 
emerged in the quasihydraulic Mycenaean period. 1 During the later 
part of the first millennium B.C., when Alexander and his successors 
ruled the entire Near East, the architectural concepts of Hellas trans- 
formed and refined the hydraulic style without, however, destroying 
its monumental quality. 

In Islamic architecture the two styles blended to create a third. 
The products of this development were as spectacular in the western- 
most outpost of Islamic culture — Moorish Spain — as they were in the 
great eastern centers: Cairo, Baghdad, Bukhara, Samarkand, and Is- 
tanbul. The Taj Mahal of Agra and kindred buildings show the same 
forces at work in India, a subcontinent which, before the Islamic in- 
vasion, had evolved a rich monumental architecture of its own. 

c. The Institutional Meaning 

It hardly needs to be said that other agrarian civilizations also com- 
bined architectural beauty with magnitude. But the hydraulic rulers 
differed from the secular and priestly lords of the ancient and me- 
dieval West, first because their constructional operations penetrated 
more spheres of life, and second because control over the entire 
country's labor power and material enabled them to attain much 
more monumental results. 

The scattered operations of rainfall farming did not involve the 
establishment of national patterns of cooperation, as did hydraulic 
agriculture. The many manorial centers of Europe's knighthood so- 
ciety gave rise to as many fortified residences (castles); and their size 
was limited by the number of the attached serfs. The king, being 
little more than the most important feudal lord, had to build his 
castles with whatever labor force his personal domain provided. 

The concentration of revenue in the regional or territorial centers 

CHAPTER 2, F 45 

of ecclesiastical authority permitted the creation of the largest in- 
dividual medieval edifices: churches, abbeys, and cathedrals. It may 
be noted that these buildings were erected by an institution which, 
in contrast to all other prominent Western bodies, combined feudal 
with quasihydraulic patterns of organization and acquisition. 

With regard to social control and natural resources, however, the 
master builders of the hydraulic state had no equal in the non- 
hydraulic world. The modest Tower of London and the dispersed 
castles of Medieval Europe express the balanced baronial society of 
the Magna Carta as clearly as the huge administrative cities and 
colossal palaces, temples, and tombs of Asia, Egypt, and ancient 
America express the organizational coordination and the mobiliza- 
tion potential of hydraulic economy and statecraft. 


1. A Comparative View 

A government capable of handling all major hydraulic and non- 
hydraulic construction may, if it desires, play a leading role also in 
the nonconstructional branches of industry. There are "feeding" in- 
dustries, such as mining, quarrying, salt gathering, etc.; and there are 
finishing industries, such as the manufacture of weapons, textiles, 
chariots, furniture, etc. Insofar as the activities in these two spheres 
proceeded on a large scale, they were for the most part either directly 
managed or monopolistically controlled by the hydraulic govern- 
ments. Under the conditions of Pharaonic Egypt and Inca Peru, di- 
rect management prevailed. Under more differentiated social con- 
ditions, the government tended to leave part of mining, salt gather- 
ing, etc. to heavily taxed and carefully supervised entrepreneurs, 
while it continued to manage directly most of the large manufactur- 
ing workshops. 

By combining these facts with what we know of the hydraulic and 
nonhydraulic constructional operations of the state, we may in the 
following table indicate the managerial position of the hydraulic 
state both in agriculture and industry. For purposes of comparison, 
we include corresponding data from two other agrarian societies and 
from mercantilist Europe. 

a. For another peculiarity of hydraulic architecture, the "introvert" character of 
most of the residential buildings, with the exception of those of the ruler, see below, 
p. 86, n. b. 


Table 1. Government Management in the Spheres of Agriculture and Industry 








Waterworks Farming 

Mining, etc. 




Hydraulic society 






Coastal city states of 
classical Greece 

— _ 

, — . 

Medieval Europe 

(+) f 





Mercantilist Europe 

— — 






-f- Predominant 
+ Outstandingly significant 

— Irrelevant or absent 

1. Simp! 

2. On a 
$. On a 

er conditions, 
national scale, 
manorial scale. 

( ) Trend limited or modified by factors indicated in the text 

In ancient Greece, mining was mainly in the hands o£ licensed 
businessmen. As long as the concessionaire delivered a fixed part of 
his output to the state, he enjoyed "very extensive" rights; he "was 
said to 'buy' the mine, he organized the working as he pleased, the 
ore was his, and he could cede his concession to a third party." x In 
Medieval Europe mining was also essentially left to private entre- 
preneurs, who, having obtained a concession from the royal or ter- 
ritorial authorities, proceeded independently and mostly through 
craft cooperatives. 2 The mercantilist governments of Europe operated 
some mines directly; but the majority was managed by strictly super- 
vised private owners. 3 

All these arrangements differ profoundly from the system of gov- 
ernment mining prevailing in Pharaonic Egypt and Inca Peru. Mer- 
cantilist usage resembles in form, but not in institutional substance, 
the policy pursued in certain of the more differentiated hydraulic 
societies, where government operation of some mines was combined 
with private, but government-licensed, handling of others. 4 

Except for mining, Oriental and Occidental absolutism are less 
similar in the industrial sphere than has been claimed, whereas a 
resemblance of sorts does exist between hydraulic society and feudal 
Europe. In hydraulic society, the majority of the not-too-many larger 
industrial workshops was government managed. In the mercantilist 
Occident they were, under varying forms of state supervision, pre- 
dominantly owned and run by private entrepreneurs. In the coastal 
city-states of classical Greece the government was neither equipped 
nor inclined to engage in industrial activities. The rulers of Medieval 
Europe, faced with a different situation, proceeded differently. In 

CHAPTER 2, F 47 

their manorial workshops they employed a number of serf -artisans, 
who were kept busy satisfying the needs of their masters. The feudal 
lords also summoned serf labor for the construction of "big houses" 
— castles. The similarity between this manorial system of cooperative 
work and the hydraulic pattern is evident. But again the functional 
similarity is limited by the differences in the societal setting. The 
medieval kings and barons could dispose only over the labor force 
of their own domains and estates, while the hydraulic rulers could 
draw on the unskilled and skilled labor of large territories, and ulti- 
mately on that of the whole country. 

The decisive difference, however, between hydraulic society and 
the three civilizations with which we compare it lies, insofar as in- 
dustry is concerned, in the sphere of construction. It is this sphere 
which more than any other sector of industry demonstrates the or- 
ganizational power of hydraulic society. And it is this sphere which 
achieved results never attained by any other agrarian or mercantilist 

The full institutional significance of this fact becomes apparent as 
soon as we connect it with the corresponding agrarian development. 
Government-managed heavy water works place the large-scale feeding 
apparatus of agriculture in the hands of the state. Government- 
managed construction works make the state the undisputed master of 
the most comprehensive sector of large-scale industry. In the two 
main spheres of production the state occupied an unrivaled position 
of operational leadership and organizational control. 

2. The Power of the Hydraulic State over Labor 
Greater than That of Capitalist Enterprises 

In both spheres the hydraulic state levied and controlled the needed 
labor forces by coercive methods that were invocable by a feudal 
lord only within a restricted area, and that were altogether different 
from the methods customary under capitalist conditions. The hydrau- 
lic rulers were sufficiently strong to do on a national scale what a 
feudal sovereign or lord could accomplish only within the borders 
of his domain. They compelled able-bodied commoners to work for 
them through the agency of the corvee. 

Corvee labor is forced labor. But unlike slave labor, which is de- 
manded permanently, corvee labor is conscripted on a temporary, al- 
though recurring, basis. After the corvee service is completed, the 
worker is expected to go home and continue with his own business. 

Thus the corvee laborer is freer than the slave. But he is less free 
than a wage laborer. He does not enjoy the bargaining advantages 


of the labor market, and this is the case even if the state gives him 
food (in the ancient Near East often "bread and beer") or some 
cash. In areas with a highly developed money economy the hydraulic 
government may levy a corvee tax and hire rather than conscript 
the needed labor. This was done largely in China at the close of 
the Ming dynasty and during the greater part of Ch'ing rule. 

But there as elsewhere the government arbitrarily fixed the wage. 
And it always kept the workers under quasimilitary discipline. 5 Ex- 
cept in times of open political^erisis, the hydraulic state could always 
muster the labor forces it required; and this whether the workers 
were levied or hired. It has been said that the Mogul ruler Akbar, 
"by his firman (order) could collect any number of men he liked. 
There was no limit to his massing of labourers, save the number of 
people in his Empire." 6 Mutatis mutandis, this statement is valid for 
all hydraulic civilizations. 


Thus the hydraulic state fulfilled a variety of important managerial 
functions. In most instances it maintained crucial hydraulic works, 
appearing in the agrarian sphere as the sole operator of large prepara- 
tory and protective enterprises. And usually it also controlled the 
major nonhydraulic industrial enterprises, especially large construc- 
tions. This was the case even in certain "marginal" areas, 1 where the 
hydraulic works were insignificant. 

The hydraulic state differs from the modern total managerial states 
in that it is based on agriculture and operates only part of the country's 
economy. It differs from the laissez-faire states of a private-property- 
based industrial society in that, in its core form, it fulfills crucial 
economic functions by means of commandeered (forced) labor. 

a. Social science is indebted to James Burnham for pointing to the power potential 
inherent in managerial control. The present inquiry stresses the importance of the 
general (political) organizer as compared not only to the technical specialist (see Veblen, 
1945: 441 ff.), but also to the economic manager. This, however, does not diminish the 
author's appreciation of the contribution made by Burnham through his concept of 
managerial leadership. 


state stronger than society 


The hydraulic state is a genuinely managerial state. This fact has 
far-reaching societal implications. As manager of hydraulic and other 
mammoth constructions, the hydraulic state prevents the nongovern- 
mental forces of society from crystallizing into independent bodies 
strong enough to counterbalance and control the political machine. 

The relations between the governmental and nongovernmental 
forces of society are as manifold as the patterns of society itself. All 
governments are concerned with the protection of the commonwealth 
against external enemies (through the organization of military action) 
and with the maintenance of internal order (through jurisdiction and 
policing methods of one kind or another). The extent to which a 
government executes these and other tasks depends on the way in 
which the societal order encourages, or restricts, governmental activ- 
ities on the one hand and the development of rival nongovernmental 
forces on the other. 

The nongovernmental forces aiming at social and political leader- 
ship include kin groups (particularly under primitive conditions); 
representatives of autonomous religious organizations (customary in 
certain primitive civilizations but, as the history of the Christian 
Church shows, by no means confined to them); independent or semi- 
independent leaders of military groups (such as tribal bands, armies 
of feudal lords); and owners of various forms of property (such as 
money, land, industrial equipment, and capacity to work). 

In some cases the rise of hydraulic despotism was probably con- 
tested by the heads of powerful clans or by religious groups eager to 
preserve their traditional autonomy. In others, semi-independent 
military leaders may have tried to prevent the masters of the hydrau- 
lic apparatus from attaining total control. But the rival forces lacked 
the proprietary and organizational strength that in Greek and Roman 
antiquity, as well as in Medieval Europe, bulwarked the nongovern- 
mental forces of society. In hydraulic civilizations the men of the 



government prevented the organizational consolidation of all non- 
governmental groups. Their state became "stronger than society." * 
Any organization that gives its representatives unchecked power over 
its subjects may be considered an "apparatus." In contrast to the con- 
trolled state of multicentered societies, the state of the single-centered 
hydraulic society was a veritable apparatus state. 


1. The Great Builders of Hydraulic Society — 
Great Organizers 

Superior organizational power may have different roots. In a hy- 
draulic setting the need for comprehensive organization is inherent 
in the comprehensive constructions necessitated or suggested by the 
peculiarities of the agrarian order. 

These constructions pose numerous technical problems and they 
always require large-scale organization. To say that the masters of 
hydraulic society are great builders is only another way of saying 
they are great organizers. 

2. Fundamentals of Effective Organization: 
Counting and Record Keeping 

An organizer combines disparate elements into an integrated whole. 
He may do this ex tempore if his aim is simple or passing. He must 
make more elaborate preparations if he is confronted with a perma- 
nent and difficult task. Dealing with human beings — their labor 
power, their military potential, and their capacity to pay taxes — he 
must know their number and condition. To this end he must count 
the people. And whenever he expects to draw from them frequently 
and regularly, he must preserve the results of his count either by 
memorizing them or, above the most primitive level, by utilizing 
preliterary or literary symbols. 

It is no accident that among all sedentary peoples the pioneers of 
hydraulic agriculture and statecraft were the first to develop rational 
systems of counting and writing. It is no accident either that the 
records of hydraulic society covered not only the limited areas of 
single cities or city states, of royal domains or feudal manors, but the 
towns and villages of entire nations and empires. The masters of 
hydraulic society were great builders because they were great organ- 
izers; and they were great organizers because they were great record 

CHAPTER 3, B 51 

The colored and knotted strings (quipus) by which the Incas pre- 
served the results of their frequent countings x show that the lack of 
a script constitutes no insurmountable barrier to numbering and 
registering the population. In pre-Conquest Mexico the various forms 
of land and the obligations attached were carefully depicted in cod- 
ices; and the procedures of local administrators were apparently 
based on these all-important documents. 2 

In China an elaborate system of writing and counting existed as 
early as the Yin (Shang) dynasty, that is, in the second millennium 
B.C. Under the subsequent Chou dynasty census lists were used for 
determining potential fighters and laborers and for estimating rev- 
enue and expenditures. Specific evidence testifies to a detailed system 
of counting and registering in the ruling state of Chou, 3 and we know 
that at the close of the Chou period the people were registered in the 
great northwestern country of Ch'in, 4 and also in Ch'i. In Ch'i the 
census is said to have been taken every year in the autumn. 5 It was 
in this season that people were also counted under the first long-lived 
imperial dynasty, Han. 6 Preserved bamboo records indicate that the 
Han registers follow a regular pattern. 7 The two sets of Han census 
figures contained in the official history of the period 8 are the most 
comprehensive population data to come down to us from any major 
contemporary civilization, including the Roman Empire. 

The later history of the Chinese census presents many problems 
which are far from solved. The methods and the accuracy of proce- 
dures changed greatly with time, but the government's role in the 
handling of these matters cannot be doubted. In one way or another, 
the imperial bureaucracy succeeded in keeping track of its human 
and material resources. 

The same holds true for India. The A rthashastra 9 and the Islamic 
sources 10 reveal the interest which both native and foreign rulers 
took in counting their subjects and estimating their revenues. And 
this interest was by no means academic. Megasthenes found various 
groups of officials in the Maurya empire charged with such tasks as 
measuring the fields and counting the people. 11 Numerous inscrip- 
tions throw light on surveys made during the last period of Hindu 
India. 12 

After China, we are probably best informed on the Near Eastern 
development of governmental counting and registering. The oldest 
deciphered inscriptions dealing with the economy of a Mesopotamian 
temple city contain many numerical data on land, people, agricul- 
ture, and public services. 13 In Pharaonic Egypt the people were 
counted regularly from the time of the Old Kingdom. 14 Documentary 
evidence for the connection between the census and fiscal and per- 


sonal obligations exist only for the Middle and New Kingdoms, but 
the absence of still earlier data on this point is certainly accidental. 15 
On the eve of the Hellenistic period persons and property seem to 
have been listed annually; 16 and the Ptolemies probably perpetuated 
the ancient system. The papyri suggest that there were two cadasters 
used for mutual checking, one in the individual villages and one in 
the metropolis. 17 

Under the succeeding regimes the methods of counting people and 
property, particularly land, underwent many modifications; but as 
in India and China the underlying principle continued to receive 
recognition. The Romans inherited the Hellenistic pattern 18 and the 
Arabs based their system on that of Eastern Rome. 19 The Mamluks 
upheld the time-honored system of record keeping, 20 as did the Otto- 
man Turks, who during the heyday of their power insisted that 
"every thirty years a census must be taken, the dead and the ill must 
be separated off, and those not on the rolls must be newly re- 
corded." 21 

3. Organizational and Hydraulic Management 

A glance at the metropolitan and local centers of hydraulic record 
keeping recalls the original meaning of the term "bureau-cracy" : 
"rule through bureaus." The power of the agromanagerial regime 
was indeed closely interlinked with the "bureaucratic" control which 
the government exerted over its subjects. 

a. The Organizational Task Inherent in Large Constructions, 
Hydraulic and Otherwise 

As stated above, enormous organizational tasks are inherent in the 
large constructions which the agrarian apparatus state accomplishes 
and which, particularly in their hydraulic form, play a decisive role 
in crystallizing the over-all conformation. Having, in the preceding 
chapter, dealt at some length with the constructional developments of 
hydraulic society, we shall confine ourselves here to re-emphasizing 
once more the cardinal importance of organization in this field. 

b. Hydraulic Management 

The outstanding forms of hydraulic management (as juxtaposed to 
construction) are the distribution of irrigation water and flood 
watching. In general, these two operations require much less man- 
power than does the work of construction and repair, but those en- 
gaged in the former must cooperate very precisely. 

CHAPTER 3, B 53 

Megasthenes describes the care with which officials of the Maurya 
empire opened and closed the canals and conduits to regulate the dis- 
tribution of the irrigation water.* The highly systematized handbook 
of Chinese statecraft, the Chou Li, speaks of special officials who con- 
ducted the irrigation water from the reservoirs and larger canals to 
the smaller canals and ditches. 22 Herodotus, in a frequently quoted 
passage, tells how in Achaemenian Persia the sovereign himself super- 
vised the major hydraulic operations: "The king orders the flood- 
gates to be opened toward the country whose need is greatest, and lets 
the soil drink until it has had enough; after which the gates on this 
side are shut, and others are unclosed for the nation which, of the 
remainder, needs it most." 23 

Megasthenes and Herodotus make it very clear that the govern- 
ment was the distributing agent of the irrigation water; but they do 
not furnish organizational details. Such data are buried in adminis- 
trative manuals and regulations which, because of their predomi- 
nantly technical nature, have received little scholarly attention. 
Among the exceptions are some accounts of 10th- and 16th- (or 
17th-) century Persia and several irrigation codes discovered in Bali. 

The documents dealing with Persian conditions show the care with 
which the available water was assigned. They indicate also the clock- 
like cooperation between the "water master" (mlrab), his subordinate 
officials and aides, and the village heads. 6 The Bali data familiarize 
us with the workings of a well-integrated hydraulic order. Here the 
ruler and the minister of revenues (sedahan agong) make the key 
decisions as to when and how to flood the various local hydraulic 
units, the subak. 24 The official head of a cluster of such units super- 
vises the supply for each subak; 25 and the chief of the local unit, the 
klian subak, coordinates the individual peasants, who swear a solemn 
oath to submit to regulations while the rice fields, sawah, are being 
flooded. 26 "Thus the orderly distribution of the water among the 
various sawah-ho\ders is accomplished with extreme care, and also 
with well-based reasons. The sawah-holder cannot at any time dis- 
pose over his share of the water supply where the water is scarce. The 

a. Strabo 15. 1. 50. Smith, 1914: 132. Buddha himself is said to have settled a 
conflict between two city states over their rights to use the waters of a nearby river 
(Jdtakam, V: 219). 

b. Lambton, 1948: 589 ff. Ibid., 1938: 665 ff. The organization of the irrigation system 
in East Persia at the time of the Abbassid caliphate is described in Arab sources. 
The head of the water office in Merv had at his disposal ten thousand hands, and his 
power surpassed that of the district police chief. The storage dam below the city 
was operated by four hundred guards; and the technique of measuring and distributing 
the water was minutely regulated (Mez, 1922: 423 ff.)- For the institution of the water 
master in ancient and modern South Arabia see Grohmann, 1933: 31. 


various sawah -holders, even if they belong to the same subak, must 
share the available water and must have their sawah flooded in 
sequence." 27 

The organizational operations involved in the distribution of the 
irrigation water are remarkable for their subtlety and for their cen- 
tralization of leadership. Conflicts are frequent between cultivator 
and cultivator and between subak and subak. "If each sawah-holdev 
could do as he pleases, there would soon arise the greatest disorder 
and the lower subak would probably never get their water." All these 
problems are successfully resolved because essentially "the distribu- 
tion of the water as well as the water law lies in the hands of a single 
person." 28 

The control of flood water necessitates greater organizational effort 
only under special circumstances. An operational problem arises 
primarily where the seasonal overflow of an extended source of water 
threatens the irrigation system and the safety of those depending on it. 
In Bali the upper courses of the river have to be watched; and espe- 
cially assigned men fulfill this function as a regular part of their 
hydraulic corvee. 28 In imperial China, even in times of decay, the 
government placed thousands of persons along their extended em- 
bankments in the battle against potential floods. 30 Between 1 883 and 
1888 the Egyptian government levied about one hundred thousand 
corviable persons annually to watch and fight the flood. 


4. The Organization of Quick Locomotion and 


Under hydraulic conditions of agriculture, certain large operations 
of construction and management must be organized. Other organiza- 
tional activities are not imperative, but they are made possible by a 
political economy which compels the government to maintain centers 
of direction and coordination in all major regions of production. 
Being able to establish its authority not only over a limited "royal 
domain" and a number of royal towns — as does the typical feudal 
state — the hydraulic regime places its administrators and officers in 
all major settlements, which virtually everywhere assume the char- 
acter of government-controlled administrative and garrison towns. 

Effective governmental control involves first the political and fiscal 
superiority of the directing agency and second the means for convey- 
ing commands and commanders to the subcenters of control. The 
desire to exert power through the control of communications char- 
acterizes all political hierarchies; but circumstances determine the 
extent to which this desire will be satisfied. The overlord of a feudal 

CHAPTER 3, B 55 

society valued fast communications as much as any Oriental despot; 
but the spotty distribution of his administrative centers and the po- 
litically conditioned lack of good roads prevented his messages from 
traveling as quickly or as safely as did the messages of the hydraulic 

The development of long highways and navigation canals is only 
another manifestation of the extraordinary construction potential of 
hydraulic society. Similarly the development of effective systems of 
communication is only another manifestation of its extraordinary 
organizational potential. Almost all hydraulic states bulwarked their 
power by elaborate systems of "postal" communication and intelli- 

The terms "post" or "postal service" express the fact that persons 
are "posted" at intervals along the road; the formula "relay system" 
points to the regulated interaction between the persons so posted. 
The terms will be used interchangeably and with the understanding 
that, within our context, they refer to an organization maintained by 
the state for the purposes of the state. On occasion the post handled 
rare and perishable goods (fruit and fish for the court, etc.). But its 
primary aim was the movement of persons of privilege (envoys, 
officials, foreign diplomats), messengers, and messages — these latter 
including intelligence of the most confidential, important, and deli- 
cate nature. 

In the decentralized society of Medieval Europe individuals or 
groups of individuals (merchants, butchers, towns) established over- 
land communications long before the government undertook the 
organization of a systematic postal service. 32 In the hydraulic world, 
private communications were not lacking, 33 but they never competed 
with the far-flung and effective relay system of the state. By running 
the post as a political institution, the representatives of Oriental 
government maintained a monopoly over fast locomotion, which — 
interlocked with an elaborate system of intelligence — became a for- 
midable weapon of social control. 

The hydraulic countries of ancient America present the relay 
system in a simple but highly effective form. In the absence of suit- 
able transport animals, messages were carried by runners, who in the 
Mexican area proceeded along more or less informal routes and in 
the Andean area, along excellent state highways. The Mexican relay 
stations are said to have been set something like two leagues (ca. 
6 miles) apart; 34 and, according to Torquemada, the speed with 
which messages could be delivered exceeded one hundred leagues 
(300 miles) per day. 35 The stations along the Inca road were closer 
to each other, at times no more than three-quarters of a mile separat- 


ing them. The runners could move at a speed of one hundred and 
fifty miles per day. According to Cobo, one message was carried from 
the coastal town of Lima to Cuzco, the capital of the altiplano, over 
approximately four hundred miles of difficult and often steep terrain, 
in something like three days. A hundred years after the conquest it 
took the Spanish horse-mail twelve to thirteen days to cover the same 
ground. While on service, the runners had to be fed; and this was 
the responsibility of the settlements through which the relay routes 
passed. 36 As a matter of fact, in all parts of the hydraulic world those 
who lived along the post roads were generally compelled to provision 
the stations, furnish auxiliary labor, and supply the draft and trans- 
port animals, carriages, sedan chairs or boats demanded by the relay 

The Incas are said to have been extremely well informed about 
the remotest regions of their empire. 37 The far-flung organization of 
the postal system of Achaemenian Persia greatly impressed Herod- 
otus. 38 Private letters might also be carried, but for security reasons 
they were read by the postal officials. 39 Xenophon stressed the intelli- 
gence angle. Through the royal post the Achaemenian kings were 
able "to learn with great celerity the state of affairs at any distance." 40 

The technical peculiarities of the Roman state post have been fre- 
quently described. The layout of its larger and smaller stations (man- 
stones and mutationes) and the organizational pattern of the institu- 
tion are indeed remarkable. 41 But it is important to remember that 
from the very beginning the cursus publicus was primarily aimed at 
providing the imperial center with information. 42 By establishing the 
post, Augustus laid the foundations for a comprehensive intelligence 
system. Special officials, first called frumentarii and from Diocletian 
on agentes in rebus, operated in conjunction with the technical staff. 
Their activities enormously strengthened the hold of the autocracy 
over its subjects. 43 

At the beginning of the Byzantine period the postal system is said 
to have been excellent. 44 According to Procopius, it enabled the cour- 
iers to cover in one day a distance otherwise requiring ten days. 45 
The Sassanid rulers of Persia followed the Achaemenian tradition 
both in maintaining an effective postal service and in using it essen- 
tially for the purposes of the state. 46 

It is generally claimed that the caliphs shaped their postal system 
after the Persian model. 47 This seems to be true with one important 
qualification. The Arabs, who carried with them the tradition of the 
steppe and the desert, moved on horseback or by means of camel 

c. Cobo, HNM, III: 269; Rowe, 1946: 231 ff. According to Cieza (1945: 137), a message 
was carried this distance in eight days. 



caravans. Consequently they paid little attention * s to the well-kept 
highways, which had been the glory of the Near Eastern postal serv- 
ice until the days of the Sassanids. Otherwise they were indeed eager 
to keep the state post in good condition. In the 9th century the cali- 
phate is said to have maintained over 900 relay stations. 49 

Under the caliphs the postmaster-general was often at the same 
time the head of the intelligence service. 50 An appointment decree 
of the year a.h. 315 (a.d. 927-28) states clearly that the caliph expected 
the head of the postal service to observe in detail the state of farming, 
the condition of the population, the behavior of the official judges, 
the mint, and other relevant matters. The secret reports were to deal 
separately with the various classes of functionaries, judges, police of- 
ficials, persons in charge of the taxes, etc. 51 The directives imply elab- 
orate methods of gathering and tabulating information. 

The Fatimids perpetuated the postal tradition of their Arab pred- 
ecessors; 62 and the Mamluks were at least as eager to maintain the 
state post, which during the period of their prosperity connected the 
Egyptian metropolis with the various regions of Syria. 53 Qalqashandi 
notes the connection between the regular postal system and the or- 
ganization of intelligence and espionage. Government offices dealing 
with these matters were under the same ministry, the Diwan of Cor- 
respondence. 54 The dispatch-bearers of the Ottoman government 
carried the regime's political and administrative correspondence 
"through the length and breadth of the Ottoman Empire." 55 

Megasthenes mentions the activities of intelligence officials in 
Maurya India; S6 and the Arthashastra and the Book of Manu discuss 
in some detail the methods to be employed by spies. 57 The relation 
between the government-maintained courier system and secret intel- 
ligence becomes clearly apparent in texts dealing with the Gupta 
period (3d-8th century a.d.); 58 and it can also be documented for 
the Muslim period. 59 In Mogul times local intelligence was bureau- 
cratically organized under an official designated as kotwdl. 00 It seems 
legitimate to assume that the national intelligence service was inter- 
linked with the road system, whose public inns (sarais) and other 
conveniences were organized "in accordance with the practice of the 
best Hindu kings in ancient times." 61 

In China the relay system developed together with state roads and 
man-made waterways. Perpetuating and elaborating earlier patterns, 62 
the masters of the empire established a postal service which, with 
numerous disruptions and modifications, lasted for more than two 
thousand years. The imperial post provided the government with 
quick and confidential information on all parts of the country. Dur- 
ing the Han period, rebellious barbarians not infrequently burned 


the postal stations. 63 A high dignitary, titled King of Yen, who con- 
spired to become emperor, set up a relay system of his own for the 
speedy transmission of messages. 04 A former official, wanted by the 
government, stated in a plaintive memorandum that the government 
began its search for him by dispatching "messages by the post service 
and the post-horse system to make a proclamation near and far." His 
pursuers "examined every footprint of man" and "followed every rut 
of the carriage." Eventually the net that was "spread ail over the em- 
pire" closed in upon the fugitive; he was caught and delivered to his 
death. 65 

The relay system of the T'ang government (618-907) operated 
through more than 1,500 stations, of which nearly 1,300 served over- 
land communications, 260 functioned as "water posts," and 86 as 
both. 66 The Liao post was also exclusively reserved for the use of 
the state; its support remained the burden of the people. "Every 
county was supposed to have its own relay stations for which the 
local population had to provide the necessary horses and oxen." 67 

Viewed against such historical precedents, Marco Polo's report of 
the postal system of Mongol China does not seem unreasonable, par- 
ticularly if we remember that the Great Khan's empire included 
many a "roadless tract." 68 The Mongol rulers of China kept an un- 
usually large number of horses. But it is noteworthy that in addition 
to maintaining many major "horse post houses," even these mounted 
conquerors had many smaller stations for the use of foot runners. 
Through the runners, whose number was "immense," the Mongol 
Empire received "despatches with news from places ten days' journey 
off in one day and night." 69 

The use of foot runners — as a supplement to the horse- and boat- 
post — continued until the last imperial dynasty, Ch'ing (1616-1912). 
In 1825 the postal service operated an elaborate network of trunk 
and branch roads with more than 2,000 express stations and almost 
15,000 stations for foot messengers. For the former the administra- 
tion budgeted 30,526 horses and 71,279 service men and for the lat- 
ter, 47,435 foot messengers. These figures cover only the technical 
personnel. Official information and secret intelligence were handled 
by regional and local officials, whose vigilance was sharpened by 
threats of severe punishment. 

The organizational effort involved in maintaining this gigantic 
network is obvious. The extraordinary opportunities for speedy and 
confidential information are no less striking. The metropolitan prov- 
ince, Chihli, alone had 185 express stations and 923 foot dispatch 
posts. Corresponding figures for Shantung are 139 and 1,062; for 
Shansi, 127 and 988; for Shensi, 148 and 534; for Szechwan, 66 and 

CHAPTER 3, B 59 

1,409; for Yunnan, 76 and 425. During the 17th and 18th centuries 
the Ch'ing government allocated as much as 10 per cent of its total 
expenditures for the maintenance of its postal system. 


5. The Organizational Pattern of Warfare in 
Hydraulic Society 

Organized control over the bulk of the population in times of 
peace gives the government extraordinary opportunities for coordi- 
nated mass action also in times of war. This becomes manifest as 
soon as we contemplate such crucial aspects of defense as the monop- 
olization and coordination of military operations, organization of 
supplies, military theory, and potential size of the armed forces. A 
comparative view of these and related features reveals the institu- 
tional peculiarities of hydraulic society in this field as in others. 

a. Monopolization and Coordination 

The sovereign of a feudal country did not possess a monopoly of 
military action. As a rule, he could mobilize his vassals for a limited 
period only, at first perhaps for three months and later for forty 
days, the holders of small fiefs often serving only for twenty or ten 
days, or even less. 71 This temporary levy tended to affect only part 
of the vassals' military strength, perhaps a third or a fourth, or a still 
smaller fraction. 72 And frequently even this fraction was not obliged 
to follow the sovereign, if he campaigned abroad. 73 

The national sovereign had full control only over his own troops, 
which in accordance with the decentralized character of society con- 
stituted only a part — and often a not very large part — of the tempo- 
rarily assembled national armies. In England the Norman Conquest 
accelerated the growth of governmental power; but even here the 
royal core was slow in prevailing. In 1300 during the Carlaverock 
campaign, the king accomplished what Tout considers a maximal 
mobilization of "horse guards of the crown." At this time the "house- 
hold" element was "roughly about a quarter of the whole number 
of men-at-arms"; at best it was "nearer a third than a quarter." 74 In 
1467 the German emperor tried to gather an army of 5,217 horsemen 
and 13,285 foot soldiers for fighting against the Turks. Out of the 
aimed-at total, the emperor's own contingent was expected to com- 
prise 300 horsemen and 700 foot soldiers, while six electors were 
expected to contribute 320 and 740 respectively; forty-seven arch- 
bishops and bishops 721 and 1,813; twenty-one princes 735 and 
1,730; various counts and seigneurs 679 and 1,383; and seventy-nine 
towns 1,059 anc * 2,926." 


In all these respects the armies of the hydraulic state proceeded on 
an entirely different level. The soldiers were not protected by demo- 
cratic checks or feudal contracts. No matter whether they held office 
land or not, they came when they were summoned; they marched 
where they were told; they fought as long as their ruler wanted them 
to fight; and there was no question as to who gave the orders or who 

The constant rotation of the many armed contingents that in ac- 
cordance with the feudal contract served only for a short period con- 
stituted a major reason for the restlessness that characterized virtually 
all compound feudal armies. Another reason was the lack of a defi- 
nite authority. Where the sovereign was little more than the first 
among equals, and where the many lords proudly insisted on the 
privileges of their position, argument easily replaced obedience. Con- 
sequently military action was marked as much by the lack of disci- 
pline as by individual valor. 


b. Training and Morale 

The army of a hydraulic state might include among those it drafted 
many persons of poor training and little fighting spirit. With regard 
to skill these men might compare unfavorably with a feudal host, 
whose members were carefully trained, and with regard to morale 
they might be inferior to the warriors of both ancient Greece and 
feudal Europe. But in planned coordination they approached the 
ancient Greeks; and they far surpassed the European chevaliers. 

Table 2. Types of Societies and Types of Fighters 





Hydraulic Society 

Classical Greece 

Feudal Europe 



t t \ 

Professional Drafted men: 
troops "militia" 
Training 4- — 
Spirit -f — 
Coordination -f- -}- 


-\- Feature developed 

— Feature weak or absent 





The Greeks, who recognized the high quality of the Oriental elite 
warriors, d commented contemptuously on the poorly trained mass of 
auxiliary soldiers, 77 who obviously were draftees. Most of them did 
indeed lack the spirited integration which was the pride of the 

d. See Herodotus' account of the conversation between the exiled Spartan king, 
Demaratus, and Xerxes (Herodotus 7. 103 f.). 

CHAPTER 3, B 6l 

Greek citizen armies. 78 But opposed to the disorderly hosts of 
Medieval Europe the well-coordinated troops of the Eastern mon- 
archies made formidable enemies. About a.d. goo the author of the 
Tactica, Emperor Leo VI, e advised his generals to "take advantage 
of their [the Franks' and Lombards'] indiscipline and disorder." 
"They have neither organisation nor drill" and therefore, "whether 
fighting on foot or on horseback, they charge in dense, unwieldy 
masses, which cannot manoeuvre." T9 In the organization of the 
Western armies "there is nothing to compare to our own orderly 
division into batallions and brigades." Their camping is poor, so 
they can be easily attacked during the night. "They take no care 
about their commissariat." Under privation, their ranks tend to 
disintegrate "for they are destitute of all respect for their com- 
manders, — one noble thinks himself as good as another, — -and they 
will deliberately disobey orders when they grow discontented." 80 

This picture of "a Western army of the ninth or tenth century, 
the exact period of the development of feudal cavalry," 81 remains 
valid, with certain modifications, for the entire age of European 
feudalism. Oman describes the hosts of the Crusades as "a mixed 
multitude, with little or no organisation."" 82 "Their want of dis- 
cipline was as well marked as their proneness to plunder; deliberate 
disobedience on the part of officers was as common as carelessness 
and recklessness on the part of the rank and file. This was always 
the case in feudal armies." 83 

The modern Egyptian historian, Atiya, ascribes the victory of the 
Turks in the last major crusade to the Christians' lack of "unity of 
arms and companies" and of "common tactics." Conversely, the 
"Turkish army was ... a perfect example of the most stringent 
discipline, of a rigorous and even fanatic unity of purpose, of the 
concentration of supreme tactical power in the sole person of the 
Sultan." 8i 

c. Organization of Supplies 

The masters of hydraulic society applied the same organizational 
devices in the military sphere that they employed with such success 
in construction and communication. In many cases, the recruits 
for war could be as comprehensively mobilized as the recruits for 
toil. The assembled armies moved in orderly fashion, and camping 

e. For reasons indicated in the Introduction, above, our presentation includes refer- 
ences to Byzantium after the Arab conquests, to the Liao empire, to Maya society, 
and to other marginal hydraulic civilizations. The marginal areas of the hydraulic 
world are more fully discussed in Chap. 6, below. 


and scouting were often highly developed. Whenever feasible, the 
armies lived off the land; but numerous means were invoked to cope 
with possible shortages. 

The Incas had a "superb supply system." 85 The Persian king, 
Xerxes, in preparation for his invasion of Greece "laid up stores 
of provisions in many places. . . . He inquired carefully about all 
the sites, and had the stores laid up in such as were most convenient, 
causing them to be brought across from various parts of Asia and 
in various ways, some in transports and others in merchantmen." 88 
The Byzantine generals were definitely concerned with the "com- 
missariat" of their troops. 87 The Arabs and Turks, at the peak of 
their power, paid considerable attention to the supply problem, 
which was handled by methods suited to their special form of war- 
fare. 88 The history of Chinese warfare is filled with references to 
precisely this matter. 


d. Planned Warfare and Military Theory 

Feudal warfare, being unfavorable to the development of tactics 
and strategy in the proper sense of these terms, 90 also failed to develop 
military theory. Medieval chronicles contain innumerable references 
to battles, and the epics of knighthood never tire of describing 
military adventures. But they are concerned essentially with the 
prowess of individual fighters. Tactical considerations remain as 
irrelevant in literature as in reality. 

In the hydraulic world the organization of warfare was elaborately 
discussed. Military experts liked to evaluate their experiences in 
treatises on tactics and strategy/ The Arthashastra shows Maurya 
India well aware of the problems of aggression and defense. 91 The 
comprehensive Byzantine literature on warfare indicates the many 
problems posed by the empire's defense strategy. 02 

The organizational trends of Islamic warfare are significantly 
foreshadowed in a passage of the Koran which assures the love of 
Allah to those who fight for him "in ranks as though they were a 
compact building." 93 Later many Muslim writers discussed military 
questions. 94 

Yet probably no great hydraulic civilization produced a more ex- 
tensive military literature than China. Contrary to the prevailing 
notion, Chinese statesmen paid much attention to military prob- 
lems; they already did so during the period of the territorial states, 
which in this respect as in so many others followed hydraulic rather 

/. The military writings of ancient Greece reflect a similar, though differently 
rooted, interest in organized warfare. 

CHAPTER 3, B 6g 


than feudal patterns. The author of The Art of War, Sun Tzu, 
however brilliant, was not the sole great military theoretician in 
this period — Sun Ping and Wu Ch'i rate as high, 96 and many of the 
ideas Sun Tzu put forth are acknowledged to have been based on 
earlier writings. 97 

Almost every major territorial state had its own school of military 
thought. 98 But no matter how early the various concepts were first 
formulated, it was in the period of the territorial states that they 
assumed their classical shape. For very pragmatic reasons the empire 
maintained a lively interest in the problems of warfare. To mention 
but one piece of evidence, all major official histories from the T'ang 
dynasty (618-907) on included special, and often large, sections on 
military affairs. 

e. Numbers 

The masters of the hydraulic state, who monopolized coordinated 
military action, could — if they so wished — raise large armies. Their 
mobilization potential was entirely different from, and greatly 
superior to, that of feudal Europe. 

In Medieval England the Normans inherited a military order 
which, in addition to a feudal elite, contained elements of an older 
tribal levy. The conquerors succeeded in preserving and developing 
these rudiments of a national army; but even in England the feudal 
state could draw on only a part of the population. 

The armies of hydraulic civilizations were not so limited. Their 
numerical strength varied with such factors as military techniques 
(infantry warfare, chariots, and light or heavy cavalry), economic con- 
ditions (a natural or a money economy), and national composition 
(indigenous rule or submission under a conquering people). But 
potentially it was large. 

Where all soldiers fight on foot — either because suitable animals 
are lacking or because charioteering or riding are unknown skills — 
numbers tend to be important, even when different parts of the 
army are differently armed and trained. In ancient Mexico, 99 as well 
as in Inca Peru, 100 the government levied large infantry armies. 
Where charioteering or riding are practiced, foot soldiers may count 
for less and their number may decrease substantially. The rise of a 
money economy favors the recruiting of mercenaries, who may 
constitute the only major standing (cadre) army or who may serve 
along with a "noble" elite. 

And then there is conquest. Often, and especially at the beginning 
of a conquest dynasty, the alien ruler will depend on his own na- 


tionals to keep his power secure; and he will give little special 
training to his newly acquired subjects. 101 

But no matter how the armies of agrarian despotism are con- 
ditioned, the advantages of size rarely disappear altogether. The best 
armies of the advanced type are usually composite bodies. 102 

As noted above, the feudal armies of Medieval Europe were small 
units of mounted elite fighters. An army dispatched by Charles the 
Bald numbered less than five thousand warriors; and on several later 
occasions the records speak only of a couple of hundred horsemen. 103 
The international armies of the crusades were usually composed of 
a few thousand to no more than ten thousand men. 9 The Arabs had 
brilliant cadre armies of mounted fighters, which were supplemented 
by sizable units of auxiliary troops. 104 The standing armies of the 
first Umayyad caliphs are said to have numbered about sixty thou- 
sand men; and the last ruler of this dynasty is credited by Ibn 
al-Athlr with a host of 120,000 soldiers. 105 Harun al-Rashid once 
undertook a summer campaign with 135,000 regular soldiers and 
an unspecified number of volunteers. 106 

Similarly illuminating is a comparison of the armies of feudal 
Europe with those of the "Western Caliphate" of Cordoba. Accord- 
ing to Islamic sources, Moorish Spain in the 10th century dispatched 
twenty thousand horsemen on a northern campaign. Lot doubts 
this figure because, in the contemporary European context, it seems 
unbelievably large. Says he: "The whole of Europe was unable to 
levy at this epoch such a number." 107 His comment is as correct 
as it is inconclusive. The distinguished historian himself notes the 
enormous revenues collected by the Cordoban caliphate: "What a 
contrast to the Carolingian Empire or the Ottoman Empire, states 
without finance! Only the emperor of Eastern Rome, the Byzantine 
basileus, had perhaps equivalent resources." 10S In another part of his 
study he credits the early Byzantine Empire with two armies of 
eighteen thousand men each, plus an unknown number of occupa- 
tion troops in Africa and Italy 109 — that is, with a force of more, 
perhaps considerably more, than 40,000 men. In view of these facts 
there is no reason to doubt that Moorish Spain, a hydraulic country 
with a very dense population and a revenue far in excess of any 
of its European contemporaries, could put into the field a host half 

g. Lot, 1946, I: 130, 175, 201. Even at the close of the Crusades, the international 
European army that fought in 1396 at Nicopolis against the invading Turks had no 
national contingent comprising more than ten thousand warriors, except that of the 
immediately threatened Hungarians. The Hungarians are said to have levied some 
60,000 men (Atiya, 1934: 67), which would indeed have been something like a levee 
en masse. 

CHAPTER 3, B 65 

as large as the army of the Byzantine Empire, whose revenues, accord- 
ing to Lot's own statement, it easily matched. 

At the time of Achaemenian Persia, foot soldiers still constituted 
the bulk of all fighting men. Herodotus estimates that the Persian 
Great King mobilized against the Greeks about two million men, 110 
including his elite fighters, the ten thousand "Immortals." U1 Del- 
briick is certainly justified in doubting that any such large force 
was actually sent to Europe, but his argument becomes problem- 
atic to the extreme when he suggests that the invasion army num- 
bered only some five or six thousand armed men. 112 Nor is there any 
reason to reject the possibility that, within its confines, the Persian 
empire was able to raise armies of several hundred thousand men. 
Munro suggests that Herodotus misinterpreted an official Persian 
source when he estimated Persia's total armed strength at 1,800,000 
men. Munro himself assumes that Xerxes could muster 360,000 men 
and that the expeditionary force against Greece might have num- 
bered 1 80,000. h 

The size of India's earlier armies, which appears "incredible at 
first sight," 113 becomes plausible through comparison with the figures 
we have for the later phase of Muslim India. According to Greek 
sources, on the eve of the Maurya empire King Mahapadma Nanda 
is said to have had 80,000 horsemen, 200,000 foot soldiers, 8,000 
chariots, and 6,000 fighting elephants; n * and the figures given for 
Chandragupta's host are, with the exception of the cavalry, much 
larger, totaling "690,000 in all, excluding followers and attend- 
ants." 115 Data for later periods claim armies of 100,000 foot soldiers 
in the Andhra kingdom and hundreds of thousands to several million 
soldiers under the last Southern Hindu kings 116 and the great 
Muslim rulers. 117 

In ancient China elite units of charioteers fought alongside large 
detachments of foot soldiers. During the later part of the Chou 
dynasty cavalry began to supplement the chariots, but apparently 
the new composite armies were more rather than less numerous. 
On the eve of the imperial period the leading territorial states are 
said to have mobilized three and a half million foot soldiers, plus 
an undefined number of charioteers and over thirty thousand horse- 
men. 118 

The Liao empire had, in the ordus, a cadre cavalry of about fifty 

h. See Munro, 1939: 271-3. Eduard Meyer (GA, IV, Pt. 1: 5) states that Herodotus' 
description of Xerxes' army, like the list of Darius' tributes and other specific pieces 
of information, was based on authentic Persian sources. Munro (ibid., 271) feels certain 
that Herodotus' list of Xerxes' army was substantially the reproduction of "an 
official document." 


to sixty thousand fighters; and its records boast a militia of a million 
men. 110 Under the Sung dynasty (960-1279) the Chinese government 
is said to have trained — poorly, but nevertheless trained — a standing 
army of more than one million soldiers. 120 The Banners of the 
Manchu dynasty were a standing army that at least during the first 
phase constituted a highly qualified cavalry elite. At the end of the 
19th century these armies, which included Manchu, Mongol, and 
Chinese Bannermen, totaled 120,000 soldiers. In addition, the gov- 
ernment also had an essentially Chinese "Green" Army, which 
numbered some five to six hundred thousand men. 121 

/. Percentages 

While noting this, we have to remember that the hydraulic civili- 
zations that maintained large armies generally also had large popu- 
lations. Yet different external and internal conditions made for a 
wide range in the percentages of the total population included in 
the fighting forces. 

The army of late Ch'ing probably constituted less than 0.2 per 
cent of the total population. In the Han empire every able-bodied 
peasant was obliged to render both labor and defense service. Theo- 
retically this affected 40 per cent of the rural population 122 or some- 
thing like 32 per cent of the entire population. The cadre army 
of the Liao dynasty amounted to about one per cent of the popula- 
tion. The peasant militia comprised, on paper, about 20 per cent. 
Herodotus' data, as interpreted by Munro, suggest that in Achaeme- 
nian Persia out of a population of less than twenty millions 123 about 
1.8 per cent could be mobilized. Assuming that the population of 
late Chou China was as large as that of the Han empire at its best, 
namely about sixty millions (which probably it was not), the average 
mobilization potential of the great absolutist territorial states would 
have been almost 6 per cent. 

Of course, there is no evidence that in any of these cases an attempt 
was made to realize the full mobilization potential. The Sung govern- 
ment, which in the 11th century levied a million soldiers from 
almost twenty million families, that is, from almost one hundred 
million people, was actually drafting slightly more than one per 
cent of its population. 

Comparison with ancient Greece and feudal Europe is instructive. 
In an emergency all able-bodied free men of a Greek city state could 
be mobilized. During the 5th century B.C., Athens may temporarily 
have had under arms over 12 per cent of the total population, and 
something like 20 per cent of all free persons. 12 * 

CHAPTER 3, C 67 

The army that the German emperor raised in 1467 may have repre- 
sented 0.15 per cent of the total population of twelve millions, and 
Charles the Bald's above-mentioned army about 0.05 per cent of 
what is estimated to have been the population of France. 125 Thus 
the extremely low percentage for the late Ch'ing period still is 
higher than the German figure for 1467, and it is almost four times 
higher than the figure for gth-century France. The difference be- 
tween the feudal ratio and our other hydraulic percentages is 

To be sure, in Medieval Europe the feudal lords, monasteries, and 
burgher towns had many more soldiers; but these soldiers, being 
in excess of the agreed-upon service quota, were not obliged to 
fight in the armies of their supreme overlord. The feudal govern- 
ment was too weak to mobilize more than a fraction of the nation's 
able-bodied men; the agrodespotic regimes, like the ancient city 
states, were not so handicapped. Technical and political considera- 
tions might induce them to employ only a small percentage of 
their subjects for military purposes. But compared to feudal con- 
ditions, even relatively small armies of hydraulic states tended to be 
quantitatively impressive; and the mass armies of agromanagerial 
regimes completely exceeded both in absolute and relative terms 
the armies of comparable feudal governments. 


1. Organizational and Bureaucratic Prerequisites 

The men who direct the constructional and organizational enter- 
prises of hydraulic society can do so only on the basis of an appro- 
priately regulated income. Special modes of acquisition emerge 
therefore, together with special modes of construction and organiza- 

The acquisition of a steady and ample governmental revenue in- 
volves a variety of organizational and bureaucratic operations as 
soon as the hydraulic commonwealth outgrows local dimensions; 
and the need for such devices becomes particularly great when the 
administrative and managerial functions are fulfilled by numerous 
full-time officials. Gradually the masters of the hydraulic state 
become as much concerned with acquisitive operations as with their 
hydraulic, communicational, and defense tasks. As will be shown 
below, under certain conditions taxation and related methods of 
proprietary control may flourish together with an integrated army 
and a state post without any relevant hydraulic enterprises. 

68 a state stronger than society 

2. Labor on the Public Fields and/or the 
Land Tax 

The incipient hydraulic community may make no special arrange- 
ments for the support of its leadership. However, the consolidation 
of hydraulic conditions is generally accompanied by a tendency to 
free the chief from agricultural work in order that he may devote 
himself completely to his communal secular or religious functions. 
To this end the tribesmen cooperate on the chief's land, as they do 
on the irrigation ditches, defense works, and other communal enter- 

The Suk, who give only a fraction of their economic effort to 
hydraulic agriculture, have no public land; but in the Pueblos the 
commoners are rallied for work on the cacique's fields. 1 This is done 
largely by persuasion; but coercion is not shunned when the 
situation requires it.° In the larger communities of the Chagga 
the ruler wields more power and disposes over much land. The 
communal work involved in its cultivation is by no means light, 
but the tribesmen receive little or no compensation for doing it — 
at most some meat and a few swallows of beer at the conclusion of 
their tasks. Thus the Chagga commoner who tells his white friend, 
"For you we are working, not as in the corvee, but as on our own 
fields," 2 manifestly performs his agricultural corvee duty without 

The masters of a developed hydraulic state depend for their 
maintenance on the population's surplus labor or surplus produce, 
on the cash equivalent of such produce, or on a combination of all, 
or some, of these sources. Work on government (and temple) fields 
was regular practice in Inca Peru, Aztec Mexico, 6 and throughout 
the greater part of Chou China. The extensive temple lands of the 
Sumerian temple cities were cultivated in the main by soldier- 
peasants, who constituted the bulk of the temple personnel; but 
the communal farmers apparently delivered only a fixed part of their 
crop to the storehouses, and this they did personally and directly. 3 

a. Aitken (1930: 385) juxtaposes "the gay working parties of the Hopi" to the 
"compulsory work for the priest-chief and on the communal irrigation ditches" in the 
Rio Grande Pueblos. Significantly, the work on the chief's field was directed by the 
war chief, the chief disciplinary agent in the Pueblos (see White, 1932: 42, 45; ibid., 
1942: 97 ff. and 98, n. 10; also Parsons, 1939, II: 884, 889), and this was the case not 
only in the hydraulically more compact eastern Pueblos but in the western Pueblos 
as well. 

b. Maya commoners, like the members of the Mexican calpulli, cultivated special 
land for the "lords," the representatives of the local and central government (see 
Landa, 1938: 104). 

CHAPTER 3, C 69 

The Sumerian arrangement contrasts sharply with the coordinated 
work teams of the Inca villages * and the "thousands of pairs" that, 
according to an old Chinese ode, jointly tilled the public fields in 
early Chou times. 5 In Pharaonic Egypt the bulk of all arable land 
seems to have been assigned to individual peasants, who, after the 
harvest had been gathered, delivered part of their crop to the ap- 
propriate officials. 6 

State farms ("domains")," on which special groups of serving men 
were employed, occurred in a number of hydraulic civilizations; 
but except for pre-Conquest America and Chou China, the majority 
of all hydraulic states d seem to have preferred the land tax to corvee 
labor on large government fields. Why? 

There is no consistent correlation between the predominance of a 
natural economy and the predominance of the public land system. 
International trade and money-like means of exchange were more 
developed in Aztec Mexico than in the Old and Middle Kingdoms 
of Egypt. Possibly the absence — or presence — of agricultural labor 
animals exerted a more basic influence. Peasants who, without bene- 
fit of such animals, tilled the land with a digging stick (as they did 
in ancient Peru and Meso-America) or with a hoe (as they did in 
the greater part of Chou China), may be effectively coordinated in 
semimilitary teams, even when they work irrigated fields, whereas 
plowing teams function more effectively when permitted to operate 
as separate units on separate fields. 

Significantly, plowing with oxen spread in China during the final 
phase of the Chou dynasty 7 that witnessed the gradual abolition of 
the public field system. The peasants of Lagash, who for the most 
part seem to have worked the temple land individually, were en- 
tirely familiar with the use of agricultural labor animals. So were 
the peasants of Pharaonic Egypt and of Hindu and Muslim India. 
Thus most of the hydraulic states, in which work animals were used 
in cultivation, were maintained by the production of individual 
farmers and not by the joint effort of an agricultural corvee. 

c. State farms, sita, flourished in India during the later part of the first millennium 
B.C. (ArthafSstra, 1926: 177 ff.). These farms, however, must be distinguished from the 
Mogul khalsa, which is often referred to as the rajah's "domain." Unfortunately, the 
term "domain" has been applied both to large sectors of public land ("the king's land") 
and to limited farmlike estates. The Mogul khalsa certainly falls within the first 
category. According to Baden-Powell (1896: 198), the Mogul rulers used the term 
khalsa to designate "the whole of the lands paying revenue direct to the Treasury." 

d. Traces of public fields are reported for certain regions of India. Whether they 
reflect primitive tribal institutions, possibly of Dravidian or pre-Dravidian origin, is an 
open question (see Baden-Powell, 1896: 179, 180; ibid., 1892, I: 576 ff.; Hewitt, 1887: 
622 ff.). 



The following table indicates different forms in which a number 
of representative hydraulic governments obtained their rural reve- 

Table 3. Rural Revenue of Hydraulic Governments 



'"Public" Land Taxes 

in Kind 

Partly in Kind, 
Partly in Cash 

Tribal societies: 




(+)' + 

Ancient America: 
Inca Peru 


The Near East: 
Sumerian temple cities 
Pharaonic Egypt 
Hellenistic and Roman 
Early Byzantium 
The Arab caliphates 
Ottoman Turkey 


+ ' 





traces 4. 

Early Chou 
Late Chou 
The Imperial period (roughly) 

Documented Transition 



4. Feature developed 

— Feature undeveloped 

or absent 

/. Some. 

2. Individual responsibility 


3. Universality and Weight of the Hydraulic Tax 

The fact that work on the public fields was usually shared by all 
corviable adult males indicates the power of the hydraulic leader- 
ship to make everyone contribute to its support. The establishment 
of a money economy goes hand in hand with greater differentiations 
in property, class structure, and national revenue. But the hydraulic 
state, as the master of a huge organizational apparatus, continues to 
impose it fiscal demands on the mass of all commoners. Comparison 
shows that in this respect it was much stronger than the governments 
of other agrarian societies. 

CHAPTER 3, C 71 

In classical Athens "the dignity of the citizen could not submit to 
personal taxes." 8 When the famous city "already held the hegemony 
in Greece, she had neither regular taxes nor a treasury"; 9 and her 
national support came essentially from customs and oversea revenues. 
In republican Rome the free citizens were equally eager to keep 
public expenses low. The only major direct tax, the tributum, 
amounted to 0.1-0.3 per cent of the taxed person's property. 6 In 
both cases the nongovernmental forces of society kept the adminis- 
trative apparatus small in both personnel and budget, distinguished 
office holders receiving only an insignificant salary or none. 

The rulers of Medieval Europe supported themselves essentially 
from their personal domains, which comprised only a fraction of 
the nation's territory. The occasional or regular fees which they 
collected in their wider territory were so limited that they dem- 
onstrate the weakness rather than the strength of the sovereign's 
fiscal power. The Norman conquerors pioneered in establishing a. 
stronger state; but for reasons discussed below even they were able 
to impose taxes on all their subjects only intermittently. 10 After a 
century of struggle a mighty knighthood restricted the king's right 
to levy taxes without the consent of the "common council" to the 
three "aids," as was the custom in almost every feudal country on 
the continent. 

It is with these agrarian societies, and not with the proto-industrial 
and industrial West, that the great societies of the East must be 
compared. The masters of hydraulic agriculture spread their tax- 
collecting offices as widely as their registering and mobilizing 
agencies. All adult males were expected to toil, fight, and pay when- 
ever the state willed it. This was the rule. Exemptions had to be 
especially granted, and even when granted, they were often canceled 
either after a prescribed period or when the grantor's reign ended. 

Rural revenue was calculated in varying ways. Sometimes adult 
males, sometimes family "heads," and sometimes land units formed 
the basis for assessment. In Babylonia the land tax was collected 
even from soldiers who held service fields. 11 The government might 
demand as a general land tax 20 per cent of the annual crop. The 
same official rate is suggested also for the New Kingdom of Pharaonic 
Egypt. 12 In India during the later part of the first millennium B.C. 
it was one-twelfth, one-sixth, or one-fourth of the crop. The 
Arthashastra permits the king, in an emergency, to take up to one- 
third (instead of one-fourth) of the crop of the cultivator of good 
irrigated land. 13 Many different rate-scales are recorded for late 
Chou and imperial China. Originally the Islamic regulations made 

e. Originally taxable property was confined to land, slaves, and animals; later it 
included property of all kinds (Schiller, 1893: 196; cf. Homo, 1927: 237). 


distinctions mainly in accordance with creed; but gradually condi- 
tions became much more involved; and, of course, they differed 
widely in time and space. The many arguments about heavy taxation 
show that, under Islamic rule, the land tax was as burdensome, and 
tended to become as universal, as in other parts of the hydraulic 

A government that keeps to the official rates is considered just; 
but most governments preferred material to moral satisfaction. 
Many a sovereign went beyond the letter of the law. The clay tablets 
of Babylonia indicate that the state, which theoretically was content 
with about 10 per cent, occasionally raised the tax "to 1/5, 1/4, 1/3, 
and even one half" of the crop. 14 

Nor is this all. The payments, which appear in official lists, are in 
most cases below, and often far below, the payments which the 
tax gatherers actually extracted. Even in the most rational of all 
hydraulic states the higher echelons of the bureaucracy found it 
difficult to exert full control over their subordinates. Often the very 
effort to compel complete delivery was lacking. 

The distribution of the total tax income among the various strata 
and categories of the officialdom varied greatly. The divergencies 
are highly significant for the distribution of power within the 
bureaucracy; but they are irrelevant from the point of view of the 
state as a whole. The fiscal power of the hydraulic apparatus state 
must be measured by the total tax that the bureaucracy in its 
entirety is able to extract from the nongovernmental population in 
its entirety. Contrasted with the almost complete absence of uni- 
versal and direct taxation in the city states of ancient Greece and 
in Rome, and compared with the pathetically feeble fiscal policy 
of feudal Europe, the scope and strength of the hydraulic system 
of taxation is striking. 

4. Confiscation 

The hydraulic state, which asserts its fiscal power so effectively in 
the countryside, pursues a similar policy also toward artisans, mer- 
chants, and other owners of mobile property not protected by 
special prerogatives. The fact is so obvious that in the present 
context we shall refrain from discussing the methods invoked for 
taxing handicraft and commerce. However, another acquisitive fea- 
ture of hydraulic statecraft does deserve comment: the seizure of 
conspicuous property by outright confiscation. 

An association of free men may ask of itself whatever sacrifices 
it holds necessary for the common weal; and occasionally it may 

CHAPTER 3, C 73 

employ the weapon of confiscation against criminals or excessively 
powerful men/ But arbitrary confiscation as a general policy is 
characteristic of a genuinely absolutist regime. Having established 
unrestricted fiscal claims, such a regime can modify them at will. In 
addition, it can encroach on private property even after all regular 
and irregular taxes have been paid. 

Under simpler conditions of power and class, there is little or no 
large independent business property; and whatever confiscation 
occurs essentially hits members of the ruling group. Under more 
differentiated conditions, business wealth becomes a favorite target, 
but attacks on the property of officials do not cease. 

Large landed property is by no means immune to confiscation. But 
it is more readily accessible to taxation than are precious metals, jew- 
els, or money, which can be hidden with relative ease and which 
are indeed carefully hidden by all except the most powerful members 
of the apparatus government. The confiscatory measures of the 
hydraulic state therefore hit with particular harshness the owners 
of mobile — and concealed — property. 

The declared reasons for confiscating the property of officials and 
other members of the ruling class are almost invariably political 
or administrative. The political reasons include diplomatic blunders, 
conspiracy, and treason; the administrative, mismanagement and fis- 
cal irregularities. Serious crimes frequently lead to the wrongdoer's 
complete political and economic ruin; lesser ones to temporary or 
permanent demotion and total or partial confiscation. Businessmen 
are primarily prosecuted for tax evasion, but they too may become 
involved in a political intrigue. In the first instance they may be 
partially expropriated; in the second, they may pay with their entire 
fortune and with their life. 

Within the ruling class, conspiracies to replace the ruler or an 
important dignitary occur periodically, and particularly during times 
of insecurity and crisis. Wanton persecutions are equally frequent. 
A power center which is both accuser and judge may declare any 
activity criminal, whatever the facts. Manufactured evidence appears 
with great regularity; and legally disguised political purges are 
undertaken whenever the masters of the state apparatus deem them 

The danger of being persecuted is augmented by the fact that 
under conditions of autocratic power the majority of all officials 
and the bulk of all wealthy businessmen tend to commit acts that, 

/. For confiscation in ancient Greece, see Busolt, GS, II: nogff. The confiscations 
during the last phase of republican Rome reflect the rise of uncontrolled Orientally 
despotic power (see below, Chap. 6). 


legally speaking, are crimes, or may be so interpreted. At the court 
and/or in the administration there are always individuals or groups 
that try to promote their own interests by winning the favor of the 
ruler or other persons of high rank. The sovereign and his close 
relatives or friends, the chancellor (vizier) or other prominent mem- 
bers of the bureaucracy are all potential targets of political intrigues. 
And in an atmosphere of absolutist power, secrecy and quasicon- 
spiratorial methods appear perfectly normal. This being the case, the 
dominant center has little difficulty in pinning the label of conspiracy 
on whomever it wishes to destroy. 

To be sure, many persons who engage in such intrigues are never 
brought to book; and many others escape with minor bruises. In 
periods of prosperity and calm this is by no means rare. But 
politically phrased accusations are an essential feature of the abso- 
lutist order; and any unusual tension may spell the doom of many 
individuals or groups. 

In the administrative sphere the borderline is similarly fluid, and 
the possibilities of disaster are similarly great. Many officials have 
to make decisions regarding goods or money; and in the absence of 
rational methods of procedure and supervision, deviations from 
prescribed standards are as usual as the attempts to increase personal 
income are alluring. The classic of Hindu statecraft describes the 
almost unlimited opportunities for embezzlement offered by such 
conditions. In what amounts to a veritable catalogue, the Arthashas- 
tra mentions some forty ways in which government funds may be 
diverted. 15 The author of the Arthashastra doubts whether any per- 
son can resist so many tempting opportunities. "Just as it is im- 
possible not to taste the honey or the poison that finds itself at the 
tip of the tongue, so it is impossible for a government servant not 
to eat up, at least, a bit of the king's revenue." 16 

The wealthy businessman is equally vulnerable. Taxation being 
the prerogative of a government whose declared demands are heavy 
and whose agents tend to go beyond the official demands, the private 
men of property seek to protect themselves as best they can. They 
hide their treasure in the ground. They entrust it to friends. They 
send it abroad.* In brief, they are driven to commit acts which make 
most of them potential fiscal criminals. 

In many instances their efforts are successful, particularly when 
they are buttressed by well-placed bribes. But a technical error or a 

g. In classical India "capital wealth was hoarded, either in the house — in large 
mansions over the entrance passage . . . under the ground, in brazen jars under the 
river bank, or deposited with a friend" (C. A. F. Rhys-Davids, 1922: 219). 

CHAPTER 3, C 75 

change in the bureaucratic personnel may shatter the uneasy balance; 
and warranted accusations combined with trumped-up charges will 
initiate actions that may ruin the accused businessman economically, 
and perhaps also physically. 

In Pharaonic Egypt officials were the essential targets of confis- 
catory actions. Members of the bureaucracy who were found guilty 
of a major crime were severely punished. A demotion usually in- 
volved the loss of revenue and property, including whatever fields 
the culprit possessed either in the form of office land or as a sine- 
cure. 17 At the beginning of a new dynasty the new ruler resorted 
to such measures to consolidate his position. 18 

Disobedience to the Pharaoh, even when conspiracy was not in- 
volved, might be severely punished. A decree of the Fifth Dynasty 
threatened "any official or royal intimate or agricultural officer," 
who disregarded a certain royal order, with the confiscation of his 
"house, fields, people, and everything in his possession." The cul- 
prit himself was to be reduced to the status of a corvee laborer. 19 

The history of Chinese bureaucracy abounds with incidents of 
demotion and confiscation. When the Ch'ing emperor, Kao-tsung 
(reign-title Ch'ien-lung) died, his all-powerful minister, Ho Shen, was 
immediately arrested and "although out of respect to the memory 
of his master he was permitted to take his own life, his huge 
accumulation of silver, gold, precious stones, and other forms of 
wealth, was confiscated." 20 

The expropriation of officials for administrative and fiscal offenses 
demonstrates the vulnerability of almost all officials. Again the 
Arthashdstra neatly formulates the crux of the matter. Since every 
official who deals with the king's revenue is inevitably tempted to 
embezzle, the government must use skilled spies 21 and informers 22 
to aid in the recovery of the state's property. Crude criteria deter- 
mine whether an official is guilty or not. Whoever causes a reduction 
of the revenue "eats the king's wealth." 23 Whoever is seen enjoy- 
ing the king's possessions is guilty. 24 Whoever lives in a miserly way 
while accumulating and hoarding wealth is guilty. 25 The king may 
"squeeze them after they have drunk themselves fat, he may transfer 
them from one job to another so that they do not devour his property 
or that they may vomit up what they devoured." 26 

Of course, in all these matters discrimination is of the essence. 
The king should treat petty crimes indulgently. 27 And he should 
also be lenient when circumstances permit. Do not prosecute even 
for a serious crime, if the offender "has the support of a strong 
party"; but "he who has no such support shall be caught hold 


o£" and, the commentary adds, "be deprived of his property." 28 
These bald maxims do not even bother with an appearance of 

Confiscation may be partial or total; and it may be invoked during 
the victim's lifetime or after his death. Post-mortem expropriation 
is frequently made easy by the fact that the deceased's family is no 
longer influential. In 934 the Abbassid caliph seized the entire 
property of his deceased vizier, al-Muhallabi, squeezing money even 
from his servants, grooms, and sailors. 29 After the death of the 
mighty North Persian vizier, the as-Sahib, "his house was surrounded 
at once; the ruler searched it, found a bag with receipts for over 
150,000 dinars, which had been deposited out of town. They were 
cashed without delay, and everything contained in the house and 
treasure room was brought into the palace." 30 After the death of 
the great general, Bejkem, in 941, the caliph "sent immediately to 
the house, dug everywhere, and gathered two millions of gold and 
silver. Eventually he ordered the earth in the house to be washed, 
and this yielded a further 35,000 dirhem," but it is doubtful whether 
he found the chests of money that Bejkem had buried in the desert. 31 

Persons suspected of having defrauded the government suffered 
all manner of mistreatment. The caliph al-Qadir (991-1031) had his 
predecessor's mother severely tortured. After her resistance was 
broken, she handed over her ready cash as well as the proceeds from 
the sale of her land. 32 

The confiscation of business fortunes follows a similar pattern. As 
stated above, any prosecution could be justified politically; and 
the international connections of the big merchants made political 
accusation easy. But in the majority of cases the offense was openly 
declared to be fiscal in nature. Frequently the line between a special 
tax (for a military campaign or other emergencies) and partial con- 
fiscation is hard to draw; but whatever the pretext, the consequences 
for the victim could be grim. The Arthashastra encourages the king 
to enlarge his treasure by demanding money from rich persons 
according to the amount of their property. 33 He may squeeze such 
persons "vigorously, giving them no chance to slip away. For they 
may bring forth what others hold (for them), and sell it." 3 * 

In the case of political accusation, spies and agents could be de- 
pended upon to supply the required evidence. A middle-class 
"traitor" might be framed in several ways. An agent could commit 
a murder on a businessman's doorstep. The owner could then be 
arrested and his goods and money appropriated. 36 Or an agent could 
smuggle counterfeit money, tools for counterfeiting, or poison into 

CHAPTER 3, C 77 

the house of the potential victim, or plant a sign of allegiance to 
some other king on his property, or produce a "letter" from an 
enemy of the state. 36 Theoretically these measures were only to 
be invoked when the victim was known to be wicked; ST but along 
with other devices they are recommended in a chapter discussing 
ways for replenishing the treasury. History shows how ready the 
average despot was to use them for precisely this purpose. "Just as 
fruits are gathered from a garden as often as they become ripe, so 
revenue shall be collected as often as it becomes ripe. Collection 
of revenue or of fruits, when unripe, shall never be carried on, 
lest their source may be injured, causing immense trouble." 38 

In the Islamic world the death of a wealthy man provided the 
government with untold opportunities for decimating or liquidating 
his possessions. "Woe to him," wails an Arab text of the gth century, 
"whose father died rich! For a long time he was kept a prisoner 
in the house of misfortune, and he [the unjust official] said [to the 
son]: 'Who knows that you are his son?' And if he said: 'My neighbor 
and whoever knows me,' then they tore his mustache until he grew 
weak. And they beat and kicked him generously. And he stayed 
in closest captivity until he threw the purse before them." 39 During 
certain periods of the Abassid caliphate, "the death of a rich private 
person was a catastrophe for his whole circle, his bankers and friends 
went into hiding, objection was raised against the government's 
inspecting the testament . . . and eventually the family bought 
itself off with a major payment." 40 

To be sure, violence and plunder are not the monopoly of any 
society. But the hydraulic mode of confiscation differs in quality 
and dimension from the acts of arbitrary violence committed in 
other higher agrarian civilizations. In classical Greece it was not an 
overwhelmingly strong government but the community of propertied 
and (later also) propertyless citizens who checked a potentially over- 
powerful leader by sending him into exile and seizing his wealth. 
In Medieval Europe the rulers had only a small staff of officials, so 
small a staff indeed that intrabureaucratic struggles of the Oriental 
kind had little chance to develop. The conflicts between the feudal 
centers of power were many and often violent; but the rival forces 
fought it out more often on the battlefield than in camera. And 
those who wished to destroy their enemies by tricks preferred the 
ambush to the legal frame-up. The opportunities for using the 
first device were as numerous as those for using the second were 

As to the fate of businessmen, men of property in classical Greece 


were not plagued by heavy direct taxes; and their medieval counter- 
parts were extremely well protected against the fiscal claims of 
territorial or national overlords. Like the former, the burghers of 
the semi-independent guild cities were in no permanent danger 
of being arrested, questioned, tortured, or expropriated by the 
officials of a centralized autocracy. True, medieval trade caravans 
were held up and robbed as they moved from town to town. But 
within the confines of their walled cities the artisans and merchants 
enjoyed reasonable safety of person and possession. 

The rulers of European absolutism schemed as ruthlessly and 
killed as mercilessly as did their Eastern confreres. However, their 
power to persecute and appropriate was limited by the landed nobles, 
the Church, and the cities, whose autonomy the autocratic overlords 
could restrict, but not destroy. In addition to this, the representa- 
tives of the new central governments saw definite advantages in 
developing the newly rising capitalistic forms of mobile property. 
Emerging from an agrarian order, which they had never controlled 
or exploited in the hydraulic way, the Western autocrats readily 
protected the incipient commercial and industrial capitalists, whose 
increasing prosperity increasingly benefited their protectors. 

In contrast, the masters of hydraulic society spun their fiscal web 
firmly over their country's agrarian economy. And they were under 
no pressure to favor the urban capitalists as did the postfeudal 
Western rulers. At best, they treated what capitalist enterprise there 
was like a useful garden. At worst, they clipped and stripped the 
bushes of capital-based business to the stalk. 


1. Four Ways of Weakening Private Property 

In a number of stratified civilizations the representatives of private 
property and enterprise were sufficiently strong to check the power 
of the state. Under hydraulic conditions the state restricted the de- 
velopment of private property through fiscal, judicial, legal, and 
political measures. 

In the preceding pages we have discussed the pertinent fiscal and 
judicial methods (taxes, frame-ups, and confiscations). Before turning 
to the political aspect of the matter we must first deal with a legal 
institution which, perhaps more than any other, has caused the 
periodic fragmentation of private property: the hydraulic (Oriental) 
laws of inheritance. 

CHAPTER 3, D ^g 

2. Hydraulic Laws of Inheritance: the Principle 

Throughout the hydraulic world the bulk of a deceased person's 
property is transferred not in accordance with his will but in 
accordance with customary or written laws. These laws prescribe an 
equal, or approximately equal, division of property among the heirs, 
most frequently the sons and other close male relatives. Among 
the sons, the eldest often has special duties to fulfill. He must care 
for his mother and his younger siblings; and he may be primarily 
responsible for the religious obligations of the family. The laws take 
all this into account. But their modification does not upset the basic 
effect: the parceling out of a deceased person's estate among his 

3. The Application 

In Pharaonic Egypt the eldest son, who had important ceremonial 
tasks, received a larger share of his father's estate. But the remaining 
children also could claim a legally prescribed share of the total. 1 

The principle of more or less even division is clearly stated in the 
Babylonian code. A present made by a father during his lifetime to 
the first-born is not included in the final settlement, but "otherwise 
they [the sons] shall share equally in the goods of the paternal 
estate." 2 Assyrian law is more complicated. Again the eldest son has 
an advantage, but all other brothers are entitled to their share. 8 

In India the eldest son's originally privileged position was grad- 
ually reduced, until the difference between him and other heirs 
virtually disappeared. 4 In the Islamic world inheritance was com- 
plicated by a number of factors, among them the freedom to will up 
to one-third of an estate. But the system of "Koranic heirs" is 
definitely fragmenting: it strictly prescribes division among several 
persons. 5 The last imperial code of China reasserts what seems to have 
been regular practice during the whole period of "developed" pri- 
vate property. A family's possessions must be divided equally among 
all sons. Failure to comply was punishable by up to one hundred 
blows with a heavy stick. 6 

In Inca Peru the bulk of all land was regulated by the state and 
its local agencies. Some grants made to relatives of the ruler or 
meritorious military or civil officials might be transferred heredi- 
tarily; but the usufruct from the inherited land was subject to equal 

a. The Koran prescribes a highly intricate division of heritable property (Koran 4. 


division. 7 In Aztec Mexico the bulk of all land was occupied by 
village communities and thus barred from full transfer at the will 
of the possessor. Some land, privately held by members of the ruling 
group, was after the holder's death divided among his heirs. 8 

4. The Effect 

a. On Regulated Villages 

A law of inheritance which prescribes a periodic division of private 
property affects different groups in hydraulic society differently. 
Peasants who live in regulated village communities may divide the 
movable property of a deceased family head, but not his fields. These 
must be kept intact or, from time to time, reassigned according to 
the recognized prerogatives or needs of the members of the com- 

b. On Holders of Small Private Property 

Entirely new problems arise when the peasants own their land 
privately and freely. Scarcity of food may reduce the number of 
potential heirs, and this is an important demographic factor in all 
hydraulic societies. However, the will to live often outwits want; and 
despite periodic or perpetual shortages, the population tends to 
increase. This inevitably means smaller farms, more toil, more 
hardship, and, frequently, flight, banditry, and rebellion. 

Demographic pressures are certainly not lacking in regulated 
villages. But they are particularly serious where private landed 
property is the rule. For in such areas the impoverishment of the 
economically weaker elements is not counterbalanced, or retarded, 
by the corporate economy of the village, which prevents both in- 
dividual economic advance and collapse. 

c. On Holders of Large Private Property 

Among the wealthy property owners another factor of hydraulic 
demography becomes important: polygamy. In hydraulic civiliza- 
tions rich persons usually have several wives; and the greater their 
fortune, the larger their harem is apt to be. The possibility of having 
several sons increases proportionately. But several sons mean several 
heirs; and several heirs mean a quicker reduction of the original 
property through equal inheritance. 

Commenting on the dynamics of Chinese traditional society, two 
modern social scientists, Fei and Chang, find it "all too true" that 

CHAPTER 3, D 8l 

in this society "land breeds no land." Why? "The basic truth is that 
enrichment through exploitation of land, using the traditional 
technology, is not a practical method of accumulating wealth." 
Landed wealth tends to shrink rather than to grow; and this essen- 
tially because of the law of inheritance; "so long as the customary 
principle of equal inheritance among siblings exists, time is a strong 
disintegrative force in landholding." 9 

The Islamic law of inheritance has a similarly disintegrative effect. 
Wherever it prevails, it "must in the long run lead to the inevitable 
parceling out even of the largest property. . . ." 10 The land grants 
in the Inca empire apparently fared no better. After a few genera- 
tions the revenue received by individual heirs might shrink to in- 
significance. 11 

5. Pertinent Western Developments 

a. The Democratic City States of Ancient Greece 

The fragmentation of landed property through more or less equal 
inheritance is certainly a significant institution. But are we justified 
in considering it characteristic primarily for hydraulic civilizations? 
"The rule of dividing up an estate on succession" also operated in 
the city states of classical Greece. Consistently applied, it "split up 
the land without ceasing." 12 In the 4th century "apart from one 
exceptional case, the largest property which Attica could show . . . 
measured 300 plethra or 64 acres." Glotz adds: "This state of things 
was common to the democratic cities." 1S 

b. The United States after the War of Independence 

And then there is the fight against entail and primogeniture in the 
early days of the United States. During and immediately after the 
American Revolution the spokesmen of the young republic vigor- 
ously attacked the perpetuities, which were correctly described as 
remnants of Europe's feudal tradition. Once the law of entail was 
abolished, the colossal aristocratic landholdings quickly dissolved. 
"By about the year 1830 most of the great estates of America had 
vanished." 14 

c. A Spectacular Contrast: the Strength of Landed Property 
in Late Feudal and Postfeudal Europe 

Similar attempts at breaking the power of large landed property 
were made in Europe after the close of the feudal period. The 


governments of the new territorial and national states attacked 
entail and primogeniture through a variety of measures, statutory 
enactments prevailing on the continent and judicial reforms in 
England. 15 Resourceful protagonists of absolutism lent the struggle 
impetus and color. But in the leading countries of Western and 
Central Europe the governments were unable for a long time to 
abolish the perpetuation of big property. In France this institution 
persisted intact until the Revolution, and in a modified form until 
1849. In England and Germany it was discarded only in the 20th 
century. 16 

6. Different Social Forces Opposed to 
Proprietary Perpetuities 

a. Small and Mobile Property 

Manifestly, the perpetuation of large landed property may be 
opposed by different social forces. The Greek legislators, who, ac- 
cording to Aristotle, 17 recognized the influence of the equalization 
of property on political society, very possibly did not identify them- 
selves with one particular social group or class. But their efforts 
benefited smaller rural property 18 as well as the new forms of 
mobile (urban) property and enterprise. It stands to reason that 
the groups which profited from a weakening of big landed property 
accomplished this result through methods that became increasingly 
effective as the city states became increasingly democratized. 

In the young United States Jefferson fought for the abolishment of 
entail and primogeniture as a necessary step toward the elimination 
of "feudal and unnatural distinctions." 19 And he based his policy on 
a philosophy which distrusted commerce and industry as much as it 
trusted the independent landowning farmers. Middle and small rural 
property may not have been directly represented among those who 
wrote the Constitution; 20 but its influence was nevertheless great. 
The Revolution, which "was started by protesting merchants and 
rioting mechanics," was actually "carried to its bitter end by the 
bayonets of fighting farmers." " 

And not only this. A few decades after the Revolution the agri- 
cultural frontier prevailed so effectively over the commercial and 
banking interests of the coastal towns that it "brought about the 
declaration of hostilities against England in 1812." M It therefore 
seems legitimate to claim that it was a combination of independent 
rural (farming) and mobile urban property that brought about the 
downfall of the feudal system of entail and primogeniture in the 
United States. 

CHAPTER 3, D 83 

b. The States of Feudal and Postfeudal Europe 

The consolidation of feudal and postfeudal landed property in Eu- 
rope was challenged by a very different force. At the height of the 
conflict the attack was conducted by the representatives of the ab- 
solutist state; and the external resemblance to the Oriental version 
of the struggle makes it all the more necessary to understand the 
exact nature of what happened in the West. 

Why were the feudal lords of Europe able to buttress their landed 
property to such an extraordinary degree? Because, as indicated 
above, in the fragmented society of Medieval Europe the national 
and territorial rulers lacked the means to prevent it. Of course, the 
sovereign, the most powerful master of land and men, did exercise a 
certain public authority. 23 He claimed certain military services from 
his seigneurs, vassals, or lords; he had certain supreme judicial func- 
tions; he was expected to handle the foreign relations of his country; 
and his authority was strengthened by the fact that the bulk of his 
vassals held their fiefs only as long as they fulfilled the obligations 
mentioned in the investiture. Thus the lords were originally posses- 
sors rather than owners of their lands; and they remained so, at least 
theoretically, even after tenure became hereditary. 

This state of affairs has been frequently described. With certain 
differences — which became especially important in such countries as 
post-Conquest England — it prevailed in the greater part of Western 
and Central Europe during the formative period of feudalism. How- 
ever, the conventional picture stresses much more strongly the rela- 
tion between the feudal lord and his ruler than the relation between 
the various lords. From the point of view of proprietary develop- 
ment, the second is pivotal. 

No matter whether the baron held his fief temporarily or heredi- 
tarily, his life was centered in his own castle and not at the royal 
court; it was his detached position that determined his personal and 
social contacts. The king might claim the military services of his vas- 
sal for some few weeks; but beyond this contractually limited period 
— which might be extended if proper payments were offered 2 * — he 
was unable to control his movements. The baron or knight was free 
to use his soldiers for private feuds. He was free to engage in the 
chase, in tournaments, and in expeditions of various kinds. And most 
important, he was free to meet with lordly neighbors who, like him- 
self, were eager to promote their joint interests. 

The atomized character of the political order stimulated the asso- 
ciation of the local and regional vassals, who singly were no match 
for the sovereign but who together might successfully oppose him. In 


the race between the growth of lordly (and burgher) power on the 
one hand and royal power on the other, the rising central govern- 
ments found themselves confronted not by the scattered feudal and 
urban forces of the early days but by organized estates capable of 
defending their economic as well as their social rights. 

In England as early as the 11th century the king's tenants-in-chief 
were known as berones; originally the term connoted a group rather 
than an individual: "that word is not found in the singular." 25 But 
it was only when the government tried to check their independence 
that the barons felt the need for united action. The final section of 
the Magna Carta has been correctly called "the first royal recogni- 
tion of the baronial right collectively to coerce the king by force." 2G 
Shortly afterward, "totius Angliae nobilitas . . . took an oath each 
to the other that they would give the king no answer except a com- 
munis responsio." 27 It was in the very century in which the English 
lords incorporated themselves as an estate that they laid the founda- 
tions for the perpetuation of their lands by entail and primogeni- 
ture. 28 

On the continent the timetable and many other details differed. 
But the over-all trend was the same. Applying to their fiefs the prin- 
ciple of indivisibility — which, with the abandonment of the feudal 
form of military service, had lost its original meaning — the noble 
landholders consolidated their property in Spain, Italy, France, and 
Germany. 29 

It is worth noting that the nobles, who kept the late feudal and 
postfeudal societies balanced, owed their proprietary success partly 
to the attitude of the absolutist bureaucracy. Among the aristocratic 
members of this bureaucracy not a few felt a deep affinity for the 
landed gentry, to which they were linked by many ties. Torn by con- 
flicting proprietary and bureaucratic interests, the representatives of 
Western absolutism did not press to the extreme their organized re- 
sistance against the privileged big landowners. In consequence, there 
emerged out of the womb of feudal society one of the strongest forms 
of private property known to mankind. 

c. Hydraulic Absolutism Succeeded Where the States of 
Occidental Feudalism and Absolutism Failed 

In late feudal and postfeudal Europe the state recognized a system 
of inheritance for the landed nobles which favored one son at the 
expense of all others. And in the modern Western world the state 
by and large permitted the individual to dispose over his property at 
will. The hydraulic state gave no equivalent freedom of decision 
either to holders of mobile property or to the landowners. Its laws of 

CHAPTER 3, D 85 

inheritance insisted upon a more or less equal division of the de- 
ceased's estate, and thereby upon a periodic fragmentation of prop- 

Among primitive peoples living on an extractive economy or on 
crude agriculture, the pattern of inheritance apparently varied 
greatly; 30 thus it is unlikely that the predecessors of hydraulic soci- 
ety in their majority maintained a one-heir system of inheritance 
which the hydraulic development had to destroy. In some cases, the 
germs of a single-heir system may have had to be eradicated. Where 
no such germs existed, the hydraulic rulers made sure that efforts to 
undermine the traditional distributive pattern could get nowhere. 
They achieved their aim by a multiplicity of methods, among which 
the standardization of the fragmenting law of inheritance was only 
the most prominent one. 

In the later feudal and postleudal societies of the West the landed 
nobles were able to create the one-sided system of inheritance called 
entail and primogeniture primarily because they were armed and 
because they were nationally and politically organized. In hydraulic 
society the representatives of private property lacked the strength to 
establish similarly consolidated and strong forms of property, first 
because the governmental monopoly of armed action prevented the 
property holders from maintaining independent military forces, and 
second because the governmental network of organization (corvee, 
state post and intelligence, integrated army, and universal taxation) 
prevented the property holders from protecting their interests by 
means of an effective national organization. 

In this setting the struggle for or against the divisibility of prop- 
erty did not become a clear-cut political issue as it did in ancient 
Greece, absolutist Europe, or the United States. And in contrast to 
the areas of open conflict the hydraulic world did not favor political 
arguments which justified — or challenged — the fragmenting law of 

7. The Organizational Impotence of Hydraulic 
Property Holders 

As an armed and ubiquitously organized force, the hydraulic regime 
prevailed in the strategic seats of mobile property, the cities, as well 
as in the main sphere of immobile property, the countryside. Its 
cities were administrative and military footholds of the government; 
and the artisans and merchants had no opportunity to become seri- 
ous political rivals. Their professional associations need not have 
been directly attached to the state, but they certainly failed to create 
strong and independent centers of corporate burgher power such as 
arose in many parts of Medieval Europe. 


The countryside fared no better. The owners of land were either 
wealthy businessmen and as limited in the scope of their organiza- 
tion as were the representatives of mobile property, or — and more 
often — they were officials or priests, and a part of — or in association 
with — the nationally organized bureaucracy. This bureaucracy might 
permit its property-holding members or associates to establish local 
organizations, such as the Chinese "sash-bearers" (inadequately trans- 
lated as "gentry") and as the priests of various temples or creeds. But 
it discouraged any attempt to coordinate landed property on a na- 
tional scale and in the form of independent corporations or estates. 

The holders of family endowments (ivaqfs) in the Islamic Near 
East kept their land undivided, because these lands were destined 
ultimately to serve religious and charitable purposes. But while the 
family waqf temporarily benefited the grantee and his descendants, 
it represented neither a secure nor a free and strong form of prop- 
erty. Although less frequently singled out for confiscation, the family 
waqfs, like the other waqfs, might be seized if the state wished it. 
They were taxed; and their beneficiaries never consolidated their 
power through a nationwide political organization. 

The family waqf resembles in its announced purpose, though 
frequently not in its immediate functions, the lands held by temples 
and priests. But contrary to the religious functionaries, the holders 
of these endowments are conspicuous not for any active participa- 
tion in public life but for their rentier-like position. Temple land, 
like secular office land, was undivided; but it is indicative of the 
relation between the hydraulic state and the dominant religions that 
the landholding priests or temples did not engage in any effective 
struggle to limit the absolutist state by constitutional checks. 

Nor did the landowning members of the bureaucracy — those in 
office as well as the nonofficiating "gentry" — organize themselves into 
a national body capable of upholding their proprietary rights against 
the acquisitive and legal pressures of the state apparatus. They were 
content to use their land as a means for comfortable living, leaving 
it to those in office to organize and operate a nationally integrated 
system of political power. The Chinese general who demonstrated 
his political harmlessness by pretending to be exclusively interested 
in acquiring land 31 strikingly illustrates the political impotence 
of Oriental property, even when it is held by men of the apparatus 
itself. 6 

b. These conditions favored what may be called the introvert character of most 
residential architecture in agrobureaucratic society, as juxtaposed to the extrovert 
architecture of the corresponding type of buildings in the West. The tendency to 
hide luxurious courtyards and dwellings behind a noncommittal facade was not 

CHAPTER 3, E 87 


Similar causes led to similar results also in the field of religion. 
The hydraulic state, which permitted neither relevant independent 
military nor proprietary leadership, did not favor the rise of in- 
dependent religious power either. Nowhere in hydraulic society did 
the dominant religion place itself outside the authority of the state 
as a nationally (or internationally) integrated autonomous church. 

1. Sole, Dominant, and Secondary Religions 

A dominant religion may have no conspicuous competitors. This 
is often the case in simpler cultures, where the only relevant repre- 
sentatives of heterodox ideas and practices are sorcerers and witches. 
Here the very problem of choice is lacking; and the hydraulic leaders 
readily identify themselves with the dominant religion. 

Secondary religions usually originate and spread under relatively 
differentiated institutional conditions. Wherever such beliefs are 
given a chance to persist (non-Hindu creeds in India; Taoism and 
Buddhism in Confucian China; Christianity and Judaism under 
Islam), the rulers tend with time to identify themselves with the 
dominant doctrine. It need scarcely be asserted that in the present 
context the word "dominant" merely refers to the social and political 
aspects of the matter. It implies no religious value judgment. 
Whether the societally dominant religion is also superior in terms 
of its religious tenets is an entirely different (and legitimate) ques- 
tion, but one which does not come within the scope of the present 

2. Religious Authority Attached to the 
Hydraulic State 

a. The Hydraulic Regime — Occasionally (quasi-) Hierocratic 

In seeking to determine the relation between hydraulic power and 
the dominant religion, we must first discard a widespread miscon- 
ception. In the hydraulic world, as in other agrarian societies, 
religion plays an enormous role; and the representatives of religion 
tend to be numerous. However, the importance of an institution does 
not necessarily imply its autonomy. As explained above, the govern- 
ment-supported armies of hydraulic civilizations are usually large, 
but the same factors which make them large keep them dependent. 

confined to wealthy commoners. It also dominated the men of the apparatus — but, 
of course, not their supreme master. 


Of course, the patterns of religion cannot be equated with the 
patterns of defense. But in both cases size results essentially from 
closeness to a governmental machine, which is capable of mobilizing 
huge resources of income. 

The majority of all hydraulic civilizations are characterized by 
large and influential priesthoods. Yet it would be wrong to designate 
them as hierocratic, "ruled by priests." Many attempts have been 
made to determine the meaning of the word "priest"; and outstand- 
ing comparative sociologists, such as Max Weber, 1 have provided us 
with a wide choice of definitions for a phenomenon whose institu- 
tional borders are not easily established. 

Obviously the priest has to be qualified to carry out his religious 
tasks, which generally include the offering of sacrifices as well as 
prayers. A qualified priest may give only a fraction of his time to 
his religious duties, the greater part of it being spent to insure 
his livelihood, or he may serve professionally, that is, full time. 

If we define priestly rule as government rule by professional 
priests, then few if any of the major hydraulic states can be so 
characterized. In a number of cases the officialdom included many 
persons who were trained as priests and who, before assuming a 
government position, acted as priests. It is important to note such 
a background, because it illuminates the role of the temples in 
the ruling complex. But it is equally important to note that when 
persons with a priestly background become prominent in the govern- 
ment, they do not, as a rule, continue to spend most of their time 
fulfilling religious duties. Thus their regimes are not hierocratic in 
the narrow sense of the term, but quasihierocratic. The few hydraulic 
governments headed by qualified priests are almost all of them of 
this latter type. 

The hydraulic tribes of the Pueblo Indians are ruled by chiefs 
who play a leading part in many religious ceremonies. However, 
except for one or a few among them — often only the cacique — these 
priest-chiefs spend the bulk of their time in farming. The Pueblo 
government is therefore represented by a hierarchy of men who, 
though qualified to hold ceremonial offices, are not in their great 
majority full-time priests. 

The city states of ancient Sumer are said to have been usually 
ruled by the head priests of the leading city temples, 2 and the 
prominent courtiers and government officials, who had an important 
role in the administration of the temple estates, 3 were quite possibly 
also qualified priests. 4 But did these men, who were theologically 

a. In the history of Sumer, professional priests appear early (Deimel, 1924: 6R.; 
Falkenstein, 1936: 58; Meissner, BA, II: 52). The ancient inscriptions mention priests 

CHAPTER g, E 89 

trained, still have time to fulfill the many religious functions of a 
professional priest? Deimel assumes that the priest-kings officiated in 
the temples only on particularly solemn occasions. 4 Their sub- 
ordinates were kept equally busy by their secular duties — and equally 
restricted in their religious activities. 

The ruler's top-ranking aides, and also no doubt many of his 
lower officials, entered the political arena because they were mem- 
bers of the country's most powerful economic and military sub-units, 
the temples. The governments of the Sumerian temple cities were 
therefore quasihierocratic. But even in Sumer the power of the 
temples seems to have decreased. The reform of the priest-king, 
Urukagina, of Lagash indicates that as early as the third millennium 
b.c. leading priestly families tried to secularize the temple land; B 
and soon after Urukagina, the great kings of Akkad and Ur succeeded 
in transferring some temple lands to the royal domains. 6 During 
the subsequent Babylonian period the temples ceased to be the 
outstanding economic sector of the society, and the bulk of the 
high officials were no longer necessarily connected with the priest- 

The Babylonian pattern is much more frequent than the Su- 
merian. As a rule, the hydraulic governments were administered by 
professional officials who, though perhaps educated by priests, were 
not trained to be priests. The majority of all qualified and profes- 
sional priests remained occupied with their religious tasks, and the 
employment of individual priests in the service of the state did not 
make the government a hierocracy. 

Among the few attempts at priestly rule in a hydraulic country b 
the Twenty -first Dynasty of Pharaonic Egypt seems particularly 
worthy of note. But the usurper-founder of this dynasty, Herihor, 
who started out as a priest, held a secular government position before 
the Pharaoh made him high priest; and he was given this position 
not to strengthen but to weaken the power of the leading priest- 
hood, that of Amon. c Like the priest-kings of Sumer, the rulers of 
Pharaonic Egypt — Herihor included — obviously spent the greater 
part of their time in carrying out their governmental tasks. From 
the standpoint of ancient Egyptian history, it is significant that out 

as well as representatives of secular occupations (Schneider, 1920: 107 ff.; Deimel, 1924: 
5 ff.; Falkenstein, 1936: 58 ff.; Deimel, 1932: 444 ff.). 

b. Tibet is discussed as a marginal hydraulic society in Chap. 6, below. 

c. Kees, 1938: ioff„ 14, 16; cf. Wilson, 1951: 288 ff. Even E. Meyer (GA, II, Pt. 2: 
10 ff.), who strongly, and probably unduly, stresses the priestly background of Herihor's 
rise to power, feels that the Twenty-first Dynasty did not succeed in establishing "a real 


of the twenty-six dynasties of the Pharaonic period at best only one 
can be classed as quasihierocratic. 

b. The Hydraulic Regime — Frequently Theocratic 

The constructional, organizational, and acquisitive activities of hy- 
draulic society tend to concentrate all authority in a directing center: 
the central government and ultimately the head of this government, 
the ruler. From the dawn of hydraulic civilization it was upon this 
center that the magic powers of the commonwealth tended to con- 
verge. The bulk of all religious ceremonies may be performed by a 
specialized priesthood, which frequently enjoys considerable free- 
dom. But in many hydraulic societies the supreme representative of 
secular authority is also the embodiment of supreme religious au- 

Appearing as either a god or a descendant of a god, or as high 
priest, such a person is indeed a theocratic (divine) or quasitheocratic 
(pontifical) ruler. Obviously, the theocratic regime need be neither 
hierocratic nor quasihierocratic. Even if the divine or pontifical 
sovereign was trained as a priest, the majority of his officials would 
not necessarily have to be so qualified. 

The chieftains of the Pueblo Indians and the Chagga, who are 
the high priests of their respective communities, occupy a theocratic 
position; and the divine quality of the Hawaiian kings is beyond 
doubt. However, under primitive agrarian conditions religious and 
secular authority are often closely combined, whether cultivation 
is carried out by means of irrigation or not. 

In contrast to the wide distribution of theocratic institutions 
among primitive agrarian peoples, theocracy developed unevenly 
in the higher agrarian civilizations. Theocratic or quasitheocratic 
trends prevailed in many state-centered hydraulic societies, whereas 
they came to nothing in ancient Greece and Medieval Europe. 

In Homeric Greece the king was of divine origin, 7 and his pre- 
eminence in religious matters was so strong that he has been called 
the "chief priest." 8 Subsequent democratic developments did not 
destroy the relation between state and religion; but they placed the 
control of both types of activities in the hands of the citizens. Strictly 
supervised by the citizen community, the state religion of ancient 
Greece developed neither a clerical hierarchy B nor a closed priestly 
order. 10 As a rule, those destined to officiate as priests were chosen 
by either lot or election. 11 Hence they lacked the training which 
plays so great a role in professional and self-perpetuating priest- 
hoods. The finances of the temples were strictly controlled by politi- 

CHAPTER 3, E gi 

cal authorities, who in their majority were similarly chosen. More- 
over, governmental leaders were not considered divine, nor did they 
act as high priests or heads of any coordinated religious order. The 
designation "theocracy," which may be applied to the primitive con- 
ditions of early Greece, therefore hardly fits the "serving" citizen 
state of the democratic period. 

In the great agrarian civilizations of Medieval Europe, nontheo- 
cratic development went still further. Attempts by Pepin and Char- 
lemagne to establish theocratic authority 12 were unable to reverse 
the trend toward feudal decentralization. Among the many secondary 
centers of proprietary, military, and political power, which restricted 
the authority of the national and territorial rulers, the Church 
proved eminently effective, since a unified doctrine and an 
increasingly unified leadership endowed its quasifeudal local units 
with quasi-Oriental organizational strength. After a prolonged period 
of intense conflict, the Church gained full autonomy. In the 11th 
century the French crown "had given way to the Holy See," 13 and 
the German Emperor Henry IV humiliated himself before Pope 
Gregory VII. For some time the struggle between secular and eccle- 
siastical power continued inconclusively, until Innocent III (i 198— 
1216) raised papal authority to such a peak that he could try, al- 
though without success, to subordinate the state to the leadership 
of the Church. 

Among the many manifestations of autonomous ecclesiastical be- 
havior the English instance is particularly instructive. In 1215 the 
English bishops together with the feudal lords forced King John to 
recognize, in the Magna Carta, the legitimacy of a balanced con- 
stitutional government. The Carta was " 'primarily' a concession 
made 'to God' in favour of the Anglican Church. ... By the first 
article the king granted 'the English Church should be free, enjoy 
its full rights and its liberties inviolate' and, in particular 'that 
liberty which is considered the greatest and the most necessary for 
the English Church, freedom of elections.' Article 42 concerning 
freedom to leave the kingdom involved for the clergy the extremely 
important right to go to Rome without the king's permission." 14 

The Church under the Carta was not just one of several groups 
of effectively organized feudal landowners. In its national as well 
as in its international organization it was different from, and in a 
way superior to, the corporations of the secular nobility. Further- 
more, it struggled for autonomy as a religious body with specific 
religious objectives and claims. But however crucial these peculi- 
arities were, the Church could not have checked the power of the 
political regime if it had not, at the same time, strengthened the 


proprietary and organizational forces of the secular nobility. As the 
religious sector of these forces, the Church in the agrarian society 
of Medieval Europe became an essentially independent entity. 15 
In achieving this goal, it fatefully supported the growth of the 
balanced late feudal order, which eventually gave birth to modern 
Western society. 

Thus whether originally they were theocratically ruled or not, 
the higher agrarian civilizations of the West did not evolve massive 
theocratic power structures. The city states of classical Greece pre- 
sented a nontheocratic combination of government and religion; and 
in Medieval Europe the secular and religious authorities, far from 
establishing an integrated system of Caesaro-Papism, crystallized into 
two spectacularly separate bodies. 

Hydraulic civilization moved in a radically different direction. 
Where tribal hydraulic governments were theocratically shaped, the 
original pattern usually persisted even under more complex institu- 
tional conditions. And where theocracy was lacking in prehydraulic 
times, it frequently emerged as part of the hydraulic development. 

A society which provided unique opportunities for the growth of 
the governmental machine left no room for the growth of a politi- 
cally and economically independent dominant religion. The agro- 
managerial sovereign cemented his secular position by attaching to 
himself in one form or another the symbols of supreme religious 
authority. In some instances his position is not conclusively theo- 
cratic, but this is more the exception than the rule. In the majority 
of all cases hydraulic regimes seem to have been either theocratic or 

The institutional diversity of the hydraulic world precludes a 
rigid correlation. But it seems that divine sovereigns appear prima- 
rily under less differentiated societal conditions. On a neolithic level 
of technology the Incas ruled theocratically over a simple hydraulic 
society. The supreme ("Unique," Sapa) Inca was a descendant of the 
Sun, and thus divine; 18 and in varying degrees his relatives shared 
this status. 17 The Sapa Inca performed the most solemn sacrifices, 18 
ranking ceremonially above the professional high priests, who were 
usually chosen from among his uncles or brothers. 19 His officials 
managed the distribution and cultivation of the temple land, 20 and 
they administered the storehouses of the temples as well as those of 
the secular government. 21 Thus the government, headed by a divine 
ruler, controlled both the country's secular affairs and the priest- 
hood of its dominant religion. 

The theocratic development of the Near East is evidenced by 
many literary and pictorial records. Arising without any conspicuous 

CHAPTER 3, E 93 

institutional attachment to — though not without cultural connec- 
tions with — Mesopotamia,* the state of ancient Egypt demonstrates 
the power potential of a highly concentrated and relatively simple 
hydraulic order. The Pharaoh is a god or the son of a god, 22 a great 
and good god. 23 He is the god, Horus, 24 a scion of the Sun god, Re. 25 
He derives "bodily" from his divine parent. 28 Being thus distin- 
guished, he is the given middleman between the gods and mankind. 
Lack of time prevents him from personally attending to most of his 
religious duties; 27 but he is a high priest, 28 and the priest of all 
gods. 29 About the exaltedness of his position there can be no doubt. 

Originally the temple services were performed in considerable 
part by royal officials, 30 and the temple administration was managed 
by the king's men. 31 But even after the crystallization of a sub- 
stantial professional priesthood, the state continued to have juris- 
diction over the temple revenues; and the Pharaohs appointed the 
individual priests. 32 This system of control prevailed throughout 
the Old and Middle Kingdoms, and even at the beginning of the 
New Kingdom. It disintegrated during the period of crisis and un- 
rest, which at the end of the Twentieth Dynasty e enabled a high 
priest to ascend the throne. 33 From the Twenty-second to the 
Twenty-fifth Dynasty, Egypt was ruled by Libyan and Nubian con- 
querors, but the Pharaohs' divine position persisted despite all 
political changes down to the Twenty-sixth and last dynasty. 34 

In ancient Mesopotamia society was from the dawn of written 
history more differentiated than in early Egypt. This may be the 
reason — or one of the reasons — why the divinity of the Sumerian 
kings is formulated in a relatively complicated way. In contrast to 
the Pharaoh who was "begotten by the god— corporeal ized in the 
king — and the queen," 35 the Sumerian king is in his mother's womb 
"endowed with divine qualities, first of all strength and wisdom." 3G 
After his birth he is nurtured by the gods; and enthronement and 
coronation confirm his divinization. 37 If, as Labat suggests, the 
deities recognize the king as divine only after his birth, he is not 
the divine offspring of divine parents, but rather their adopted 

son. 38 

The controversy concerning the exact nature of the king's divinity 
in ancient Mesopotamia 39 indicates the complexity of the early 
Mesopotamian pattern, but it cannot hide the fact that the Sumerian 
king, in one way or another, represented supreme divine authority 

d. Contact between the two civilizations probably began long before the dawn 
of written history (cf. Kees, 1933: 7 ff.). 

e. For the establishment of an independent temple economy during the Twentieth 
Dynasty see Breasted, 1927, IV: 242 ff.; cf. Rostovtzeff, 1941, I: 281 ff. 


on earth. 40 He held the position of high priest. 41 In principle he was 
"the only sustainer of the high priest's office." 42 His administrative 
control over the temples was easily maintained, since in the Sumerian 
city states all major temples were headed by the priest-king, his wife, 
or some other member of his family. 43 

From the end of the Sumerian period on, the relations between 
the governments of Mesopotamia and the temples grew less close, 
but the temples were unable to free themselves from the control 
of the secular ruler. The king continued to occupy a quasidivine 
position, similar to that held by his Sumerian predecessors. As of 
old, he had the right to perform the highest religious functions. In 
Assyria he did so personally, 44 whereas in Babylonia these tasks were 
usually delegated to a representative. 45 Usually, not always. In the 
great "creation" rites at the New Year he played so important a 
religious role 48 that "during these ceremonies the sovereign was for 
his people really the very incarnation of the gods." 47 

In Assyria the government maintained strict administrative and 
judicial control over the dominant religion; 48 in Babylonia control 
was much less rigid. But here, too, the kings successfully upheld 
their right to appoint the high-ranking priests, 49 and having been 
appointed by the sovereign, "the priest had to swear an oath [of 
allegiance] like all other officials." 50 

The Achaemenian kings, who through conquest made themselves 
masters of the entire Near East, are said to have lacked divinity. Did 
they retain in their Persian homeland certain of their earlier non- 
theocratic concepts? Or were they worshiped as divine beings by 
their Persian subjects, because they were imbued with a divine 
substance? 61 Whatever the answer to these questions may be, the 
victorious Cyrus adopted in Babylonia "all the elements of Chaldean 
monarchy," 52 including royal divinity; and his successors acted 
similarly in Egypt. Like all earlier Egyptian rulers known to us, 
Darius was called divine: "Horus" and the "good god." 53 

The Hellenistic sovereigns of the Ptolemaic and Seleucid empires 
quickly learned to combine religious and secular authority. 54 Signifi- 
cantly the worship of the king was less fully developed at the in- 
stitutional fringe of the hydraulic world, in Anatolia. But here, too, 
the Hellenistic rulers definitely, if cautiously, sought theocratic 
status. 55 

The Romans adopted many of the institutions of their new 
Oriental possessions. Acceptance of the emperor's divinity was 
gradual; but the beginnings of emperor worship go back to the early 
days of the empire. The cult, which had already been proposed by 
Caesar, 56 was officially established by the first emperor, Augustus." 


CHAPTER 3, E 95 

In Early Byzantium, Christianity adjusted itself to an autocratic 
regime that felt "completely competent to legislate in all religious 
as in all secular affairs"; 58 but it proved incompatible with the con- 
cept of a divine ruler. Despite significant efforts to assert the quasi- 
divine quality of the emperor, 59 the Byzantine government was, 
according to our criteria, at best marginally theocratic. 

Islam objects to the divinization of the ruler for reasons of its 
own: Mohammad was Allah's prophet, not his son; and the caliph, 
who inherited the prophet's authority, had no divine status. Al- 
though he was in charge of important religious matters, 80 he cannot 
well be called a high priest either. Measuring the position of the 
caliph by our criteria, we therefore, and in conformity with expert 
opinion, consider it neither theocratic nor hierocratic.' 

In China the ruler emerges in the light of history as the supreme 
authority both in secular and religious matters. Whether the tradi- 
tional designation, "Son of Heaven," reflects an earlier belief in the 
sovereign's divinity, we do not know. The overlords of the Chou 
empire and of the subsequent imperial dynasties, who all used this 
appellation, were considered humans, yet they occupied a quasi- 
theocratic position. Entrusted with the Mandate of Heaven, they 
controlled the magic relations with the forces of nature by elaborate 
sacrifices. In the great religious ceremonies the ruler and his central 
and local officials assumed the leading roles, leaving only secondary 
functions to the professional sacerdotalists and their aides. The 
emperor was the chief performer in the most sacred of all ceremonies, 
the sacrifice to Heaven; 61 and he was the chief performer also in 
the sacrifices to Earth, for the prospering of the crop, 62 for the 
early summer rains, 83 and for the national deities of Soil and Millet. 64 
Some of these rites were confined to the national capital. Others 
were also enacted in the many regional and local subcenters of state 
power by distinguished provincial, district, or community officials: 
the great rain sacrifice, 65 the ceremonial plowing, 68 the sacrifices to 
Confucius 67 and to the patron of agriculture, 68 etc. ff 

To sum up: in the Chinese state religion, the ruler and a hierarchy 
of high officials fulfilled crucial priestly functions, although in their 

/. See Arnold, 1924: 189 ff., 198 n.; ibid., 1941: 294. All this is true essentially for 
the Sunnite sector of the Islamic world. In the Shi'ite sector the theocratic tendencies 
occasionally became very strong. For instance, Shah Isma'il of the Safawid Dynasty 
apparently "considered himself as God incarnate" (Minorsky, 1943: 12 n.). 

g. Thus in the political order of traditional China religious ideas and practices 
played a significant role, and certain of the latter were as comprehensive as they were 
awe-inspiring. The outstanding European expert on Chinese religion, De Groot, calls 
the great sacrifice to Heaven "perhaps the most impressive ceremony ever performed 
on earth by man" (De Groot, 1918: 180). 


vast majority these officials and the emperor himself were primarily 
occupied with secular matters. The government of traditional China 
therefore presents a consistent — and unusual — variant of theocracy. 

c. Agrarian Despotism Always Keeps the Dominant Religion 
Integrated in Its Power System 

Thus within the hydraulic world some countries were ruled quasi- 
hierocratically by qualified priests who, however, no longer engaged 
professionally in their vocation; and many were ruled theocratically, 
or quasitheocratically, by divine or pontifical sovereigns. Of the re- 
mainder some were borderline cases; and others were probably 
neither hierocratic nor theocratic. But even among the latter the 
dominant religion was unable to establish itself as an independent 
church vis-a-vis the government. In one form or another, it became 
integrated in the power system of the hydraulic regime. 

In certain regions of pre-Conquest Mexico the political ruler was 
originally also the supreme priest, 69 and in Michoacan this pattern 
persisted until the arrival of the Spaniards. 70 In the territorial states 
on the Lake of Mexico the two functions were manifestly separated 
long before the conquest, but the king continued to fulfill certain 
religious tasks, and the temples and their personnel were under his 
authority. On occasion the sovereign, alone or together with his 
top-ranking aides, might don priestly attire; 71 and he personally 
performed certain sacrifices. 72 Furthermore, and perhaps most im- 
portant, the king and his top-ranking aides appointed the Great 
Priests; 73 and temple land was apparently administered together 
with government land. 74 

Should we for this reason call pre-Conquest Mexico quasitheo- 
cratic? Perhaps. The Mexican constellation defies simple classifica- 
tion, but this much is certain: The priests of the various temples 
who assembled for ceremonial purposes had no independent nation- 
wide organization of their own. Cooperating closely with the secular 
leaders, whose offspring they educated and in whose armies they 
served, 75 they were no counterweight to, but an integral part of, 
the despotic regime. 

The borderline cases of early Achaemenian Persia and of By 
zantine and Islamic society have already been touched upon. But 
even when in these cases the government was only peripherally 
theocratic, the dominant religion was everywhere firmly enmeshed 
in the secular system of authority. The Achaemenian king, who in 
secular matters ruled absolutely, in theory also had the final say in 
religious matters. And not only in theory. The case of Artaxerxes II 

CHAPTER 3, E gy 

shows that the Achaemenian king could change the religious cult in 
significant ways. 78 The dominant priests, the magi, constituted a 
privileged group, 77 but they did not establish a national and auton- 
omous Church. 

Early Byzantium is among the very few hydraulic civilizations 
that permitted the dominant religion to function as a Church. But 
while this Church was well organized, it did not evolve into an 
independent entity, as did the Roman branch after the collapse 
of the Western half of the empire. During the early period of By- 
zantine history — that is, from the 4th to the 7th century — the 
"saintly," 78 if not divine, emperor followed Roman tradition which 
held that the religion of his subjects was part of the jus publicum; 
he consequently exerted "an almost unlimited control over the life 
of the Church." 79 

Under Islam, political and religious leadership was originally one, 
and traces of this arrangement survived throughout the history of 
the creed. The position of the Islamic sovereign (the caliphs and 
sultans) underwent many transformations, but it never lost its 
religious quality. 80 Originally the caliphs directed the great com- 
munal prayer. Within their jurisdictions, the provincial governors 
led the ritual prayer, particularly on Fridays, and they also delivered 
the sermon, the khutba. The caliphs appointed the official inter- 
preter of the Sacred Law, the muftis The centers of Muslim wor- 
ship, the mosques, were essentially administered by persons directly 
dependent upon the sovereign, such as the kadis; and the religious 
endowments, the waqfs, which provided the main support for the 
mosques, were often, though not always, administered by the govern- 
ment. Throughout the history of Islam the ruler remained the top- 
ranking authority for the affairs of the mosque. "He interfered in 
the administration and shaped it according to his will," and he 
"could also interfere in the inner affairs of the mosques, perhaps 
through his regular agencies." 82 All this did not make the caliphate 
a theocracy, but it indicates a governmental authority strong enough 
to prevent the establishment of an Islamic Church that was inde- 
pendent of the state. 

In India the relation between secular and religious authority 
underwent considerable transformation, but certain basic features 
persisted throughout and even after the close of the Hindu period. 
Available evidence suggests that in the early days of Hindu history 
the government depended less on priestly participation than it has 
since the later part of the first millennium b.c. 83 But whatever 
changes have occurred in this respect, secular and religious authority 
remained closely integrated. 


Were the Brahmins disinclined, or unable, to create an auton- 
omous position similar to that of the Church in feudal Europe? Did 
they live by gifts and government grants because they wanted to or 
because they had no choice? Everything we know about the attitudes 
of the Brahmins shows that they, like other priestly groups, preferred 
a strong and secure position over one that was weak and insecure. 
However, the Hindu sovereigns willed it otherwise. Like their 
hydraulic fellow monarchs, they favored regulated and weak forms 
of property for their subjects. They paid their secular aides in 
money, consumable goods, and the usufruct of land ("villages"); and 
they remunerated the representatives of the dominant religion in 
exactly the same way. In India this was still the policy at the end 
of the Hindu period, when an increase in private landownership 
failed to consolidate proprietary power in any way comparable to 
that of late feudal or postfeudal Europe. 

To say this does not mean to deny the extraordinary role of 
Brahminism — and of the Brahmins — in the governments of Hindu 
and Muslim India. All four castes are said to have been made from 
parts of Brahma's body, and the Brahmin caste from a particularly 
noble part, the mouth. 84 But the great Law Book ascribed to Manu 
especially stresses the divinity of the king. 80 It thus credits his rule 
with a definitely theocratic quality. 

Hindu government also had significant quasihierocratic features. 
From Vedic times the king had had a priest attached to his person, 
the purohita; 86 and this dignitary soon became his advisor in all 
matters of importance. 87 The Law Books, which were written by 
Brahmins and accepted by the government as guides for action, re- 
quire the king to have a purohita 88 "(who shall be) foremost in all 
(transactions). Let him act according to his instructions." 89 

A priest advised the king; and a priest aided him in administering 
the priest-formulated laws. The Book of Manu insists that "a learned 
Brahmana must carefully study them, and he must duly instruct his 
pupils in them, but nobody else (shall do it)." 90 In doubtful cases 
well-instructed Brahmins were to decide what was right, 91 and in the 
courts the priests, either with the king and his aides or alone, were 
to act as judges. 92 

Well educated and politically influential, the priests had unique 
opportunities for handling administrative tasks. The purohita might 
become the king's top-ranking minister. 93 In a similar way, priests 
might be entrusted with all manner of fiscal tasks. Tnis was so 
during the classical days of Hindu culture, 94 and it continued to be 
a major trend until the end of the Muslim period. Du Bois states 
that "Brahmins become necessary even to the Mussulman princes 


themselves, who cannot govern without their assistance. The Mo- 
hamedan rulers generally make a Brahmin their secretary of state, 
through whose hands all the state correspondence must pass. Brah- 
mins also frequently fill the positions of secretaries and writers to 
the governors of provinces and districts." 95 

The English did little to change this age-old pattern. The Brah- 

occupy the highest and most lucrative posts in the different 
administrative boards and Government offices, as well as in the 
judicial courts of the various districts. In fact there is no branch 
of public administration in which they have not made them- 
selves indispensable. Thus it is nearly always Brahmins who 
hold the posts of sub-collectors of revenue, writers, copyists, 
translators, treasurers, book-keepers, etc. It is especially diffi- 
cult to do without their assistance in all matters connected 
with accounts, as they have a remarkable talent for arithmetic. 
I have seen some men in the course of a few minutes work out, 
to the last fraction, long and complicated calculations, which 
would have taken the best accountants in Europe hours to get 
through. 06 

During the Hindu period and after, many trained and qualified 
priests indeed fulfilled important government functions. But except 
for the purohita and perhaps certain others who temporarily acted 
as judges, the priests became full-time officials. As in other hydraulic 
civilizations, they preserved their religious quality, but they ceased 
to be professional priests. In all probability, they did not constitute 
the majority of all officials, for there already existed a numerous 
"ruling" caste, 97 the Kshatriya, who were specialists in administrative 
and, particularly, military matters. 

d. The Changing Position of the Dominant Priesthood in 
Hydraulic Society 

These observations protect us against assuming that, during an 
early phase, hydraulic civilization was ruled by priests and that, later 
on, it was dominated by a secular group, preferably warriors. 

To repeat: hierocracy, the rule of priests who remained officiating 
priests while they governed, was rare; and rule by trained priests was 
far from being a general feature of early hydraulic civilizations. 
Theocracy characterized many hydraulic civilizations, both late and 
early; but it did not necessarily involve priest rule. 

True, in the early days of Mesopotamia and of many (most?) 


hydraulic areas of the Western hemisphere, the temples apparently 
played a dominant role in the choice of sovereigns and officials; but 
in several major hydraulic centers of the Old World this was not 
the case. In China no conspicuous body of professional priests repre- 
sented the dominant religion. In Pharaonic Egypt a professional 
priesthood was not lacking; but in the Old Kingdom many important 
religious functions were fulfilled by the ruler and certain ranking 
officials. In the early days of Aryan India the government was run 
by secular "warriors" (Kshatriyas). Only later and gradually did the 
priests, directly or indirectly, participate in the government. 

Nor can it be said that later and larger hydraulic societies were 
generally ruled by military men. As will be explained more fully in 
subsequent chapters, military officials and "the army" might indeed 
prevail over the civil bureaucracy. But this development was by no 
means confined to later and more complex hydraulic societies. More- 
over, for obvious reasons, it was the exception rather than the rule, 
since in an agromanagerial state the political organizer (the "pen") 
tends to be more powerful than the military leader (the "sword"). 


But whatever the deficiencies of this assumption of a development 
from priest rule to warrior rule, it has the merit of drawing attention 
to the multiple functions of the hydraulic regime. Different from 
the society of feudal Europe, in which the majority of all military 
leaders (the feudal barons) were but loosely and conditionally linked 
to their sovereigns, and in which the dominant religion was inde- 
pendent of the secular government, the army of hydraulic society 
was an integral part of the agromanagerial bureaucracy, and the 
dominant religion was closely attached to the state. It was this 
formidable concentration of vital functions which gave the hydraulic 
government its genuinely despotic (total) power. 



espotic power— total and not benevolent 

The despotic character of hydraulic government is not seriously 
contested. The term "Oriental despotism," which is generally used 
for the Old World variants of this phenomenon, connotes an ex- 
tremely harsh form of absolutist power. 

But those who admit the ruthlessness of Oriental despotism often 
insist that regimes of this type were limited by institutional and moral 
checks which made them bearable and at times even benevolent. 
How bearable and how benevolent was hydraulic despotism? Ob- 
viously this question can be answered only by a comparative and 
reasoned examination of the pertinent facts. 


1. Absence of Effective Constitutional Checks 

The existence of constitutional regulations does not necessarily in- 
volve the existence of a constitutionally restricted government. All 
governments that persist over time — and many others as well — have 
a certain pattern (constitution). This pattern may be expressed in 
written form. Under advanced cultural conditions, this is usually 
done, and at times in an orderly collection, a code. 

The development of a written constitution is by no means identical 
with the development of a "constitutionally" restricted government. 
Just as a law may be imposed by the government (lex data) or agreed 
upon both by governmental authority and independent nongovern- 
mental forces (lex rogata), so a constitution may also be imposed or 
agreed upon. The term constitutiones originally referred to edicts, 
rescripts, and mandates that were one-sidedly and autocratically 
issued by the Roman emperors. 

Even a highly systematized law code does not bind the autocratic 
lawgivers by restrictions other than those inherent in all self-imposed 
norms. The ruler who exercises complete administrative, managerial, 
judicial, military, and fiscal authority may use his power to make 
whatever laws he and his aides deem fit. Expediency and inertia 



favor the perpetuation of most of these laws, but the absolutist 
regime is free to alter its norms at any time; and the history of 
hydraulic civilizations testifies to the periodic promulgation of new 
laws and new codes. The "Collected Regulations" (hui yao) of im- 
perial China, 1 the Law Books (dharma shastra) of India, 2 and the 
administrative and judicial writings of the Byzantine and Islamic 
East are all cases in point. 

Having been imposed one-sidedly, constitutional regulations are 
also changed one-sidedly. In China "all legislative, executive and 
judicial powers belonged to him [the emperor]." 3 In Hindu India 
"constitutionally the king was in a position to accept or repudiate 
the laws accepted by his predecessor." * In Byzantium "there was no 
organ in the state that had a right to control him [the emperor]." 
Or, more specifically: "For his legislative and administrative acts, 
the monarch was responsible to none, except to Heaven." s 

In Islamic society the caliph, like all other believers, was expected 
to submit to the Sacred Law, 6 and generally he was quite ready to 
uphold it as part of the dominant religious order. But he asserted 
his power whenever he thought it desirable by establishing (adminis- 
trative) secular courts and by directing them through special decrees 
(qdnun or siyasa). 7 And the religious judges, the kadis, were eager 
to support a government that appointed and deposed them at will. a 
Thus the theoretical absence of a legislature modified the appear- 
ance but not the substance of Islamic absolutism. "The Caliphate 
. . . was a despotism which placed unrestricted power in the hands 
of the ruler." 8 

In these and other comparable instances the regime represents a 
definite structural and operational pattern, a "constitution." But 
this pattern is not agreed upon. It is given from above, and the rulers 
of hydraulic society create, maintain, and modify it, not as the 
controlled agents of society but as its masters. 

2. Absence of Effective Societal Checks 

a. No Independent Centers of Authority Capable of Check- 
ing the Power of the Hydraulic Regime 

Of course, the absence of formal constitutional checks does not 
necessarily imply the absence of societal forces whose interests and 

a. Schacht, 1941: 677. The Sacred Law, the Islamic law proper, was in time confined 
essentially to personal matters, such as marriage, family, and inheritance, while secular 
law dealt primarily with criminal cases, taxation, and land problems. This was so 
not only under the Arab caliphs, but also under the Turkish sultans. 

CHAPTER 4, A 103 

intentions the government must respect. In most countries of post- 
feudal Europe the absolutist regimes were restricted not so much 
by official constitutions as by the actual strength of the landed 
nobility, the Church, and the towns. In absolutist Europe all these 
nongovernmental forces were politically organized and articulate. 
They thus differed profoundly from the representatives of landed 
property, religion, or urban professions in hydraulic society. 

Some of these groups were poorly developed in the Orient, and 
none of them congealed into political bodies capable of restricting 
the hydraulic regime. The Indian scholar, K. V. Rangaswami, cor- 
rectly describes the situation when, in his discussion of Hindu ab- 
solutism, he defines genuine absolutism as "a form of government in 
which all the powers must be vested in the hands of the Ruler, there 
being no other concurrent and independent authority, habitually 
obeyed by the people as much as he is obeyed, and which lawfully 
resist him or call him to account." 9 

b. The So-called Right of Rebellion 

The lack of lawful means for resisting the government is indeed a 
significant feature of despotism. When such means are not available, 
discontented and desperate men have time and again taken up arms 
against their government, and under extreme conditions they have 
succeeded in overthrowing it altogether. Subsequently the new rul- 
ers justified their procedure by juxtaposing the worthiness of their 
cause to the un worthiness of the former regime; and the historians 
and philosophers have in the same manner explained periodic dy- 
nastic changes. It is from events and ideas of this kind that the so- 
called right of rebellion has been derived. 

The term "right of rebellion" is unfortunate in that it confuses a 
legal and a moral issue. The official discussions on the rise and fall 
of dynastic power were presented as warnings against rebellious ac- 
tion rather than as guides for it; and they were certainly not incorpo- 
rated into any official constitutional regulations or laws. The right 
of rebellion could be exercised only when the existing laws were 
violated and at the risk of total destruction for whoever asserted it. 

Traces of the so-called right of rebellion can be found in virtu- 
ally all hydraulic societies. Pueblo folklore proudly relates successful 
action against unworthy caciques, 10 and revolutions in Bali have 
been so justified. 11 Hindu and Muslim rulers have been similarly 
warned — and similarly challenged. 12 The fact that in China the right 
of rebellion was formulated in the Confucian classics did as little 
to check total power 18 as does the presence in the USSR of Marx' 


and Lenin's writings, which postulate revolutionary action against 

c. Election of the Despot — No Remedy 

Nor does the regime become less despotic because the ruler attains 
his position through election rather than through inheritance. The 
transfer of title and authority to a close relative of the deceased sov- 
ereign, preferably to the oldest son, favors political stability, while 
election favors gifted leadership. The first principle prevails among 
the indigenous rulers of hydraulic societies, the second among pas- 
toral or other peoples who, as conquerors of such societies, fre- 
quently perpetuated their original patterns of succession. 14 

The Byzantine custom of determining the emperor through elec- 
tion goes back to republican Rome. It suited the conditions of the 
early empire, which, being largely controlled by military officials, 
chose its sovereigns more often through "the army" 15 than through 
the top-ranking body of civil officials. When, from Diocletian on, the 
Senate took a more prominent part in the election of the emperor, 
the political center of gravity shifted from the military to the civil 
branch of the officialdom. 6 Election was not the best method by 
which to establish a new emperor, but wrapped in the cloak of tra- 
dition and legitimacy it proved definitely compatible with the re- 
quirements of bureaucratic absolutism. And the frequent changes 
in the person of the supreme leader deprived neither his position 
nor the bureaucratic hierarchy, which he headed, of its despotic char- 

In ancient Mexico and in most Chinese dynasties of conquest the 
new ruler was elected from members of the ruling kin group. The 
procedure combined the principle of inheritance with the principle 
of limited choice; and, as in the case of Byzantium, those who made 
the choice were top-ranking members of the political hierarchy. This 
arrangement increased the political opportunities among the masters 
of the apparatus, but it did not increase the authority of the non- 
governmental forces of society. 

Two nonhydraulic parallels may aid in dispelling the misconcep- 
tion that despotic power is democratized by an elective system of 
succession. The regime of Chingis Khan, which was perpetuated 

b. The Byzantine Senate was nothing but "the rallying-point of the administrative 
aristocracy" (Diehl, 1936: 729). 

c. Dynastic forms of government crystallized only after the Byzantine state had 
lost its hydraulic provinces. 


through limited election, remains one of the most terrifying exam- 
ples of total power. And the transfer of leadership from one member 
of the Bolshevik Politburo to another makes the Soviet government 
temporarily less stable but certainly not more democratic. 

Mommsen called the state of Eastern Rome "an autocracy tem- 
pered by a revolution which is legally recognized as permanent." 16 
Bury translates Mommsen's unwieldy formulation as "an autocracy 
tempered by the legal right of revolution." 17 Both phrasings are 
problematic because they imply that the subjects were legally en- 
titled to replace one emperor by another. Actually no such right 
existed. Diehl recognizes this by speaking of "an autocracy tempered 
by revolution and assassination"; 18 and Bury admits that "there was 
no formal process of deposing a sovran." But he adds, "the members 
of the community had the means of dethroning him, if the govern- 
ment failed to give satisfaction, by proclaiming a new emperor." 19 

This was indeed the pattern established by the military officials of 
Eastern Rome; and congruent with it, usurpation was considered 
legitimate if and when it was successful. That is, rebellion becomes 
legal — post festum. Says Bury: "If he [the pretender] had not a suffi- 
cient following to render the proclamation effective and was sup- 
pressed, he was treated as a rebel." 20 

Thus, in Byzantium as in other states of the hydraulic world, any- 
one might try to usurp power; and the elective nature of sovereignty 
combined with the temporary dominance of military leadership in- 
spired frequent attempts of this kind. But no law protected such 
actions while they were being undertaken. In Byzantium persons at- 
tacking the existing government were punished with barbarous bru- 
tality. 21 In China persons caught while trying to exercise the right of 
rebellion were executed. Under the last three dynasties they were cut 
to pieces. 22 

If armed conflict, rebellion, and the assassination of weak rulers 
do not make Oriental despotism more democratic, do they not at 
least give the populace some relief from oppression? The argument 
has less validity than may appear at first glance. Such diversions 
rarely reduce in any decisive way the traditional administrative and 
judicial pressures; and the inclination to assert supreme leadership 
through open violence is more than likely to intensify the tendency 
to brutality among those in power. Furthermore, the devastations of 
any major civil war generally lay increased economic burdens on the 
commoners. The frequent occurrence of violence within the ruling 
circles, far from tempering despotism, tends to make it more oppres- 


d. Intragovernmental Influences: Absolutism and Autocracy 

But are there perhaps forces inside the government that mitigate 
the ruthlessness of agromanagerial despotism? This question focuses 
attention on the relation between absolutism and autocracy. Absolu- 
tism and autocracy are not identical, but they interlock closely. A 
government is absolutist when its rule is not effectively checked by 
nongovernmental forces. The ruler of an absolutist regime is an auto- 
crat when his decisions are not effectively checked by intragovern- 
mental forces. 

The absolutist regimes of hydraulic society are usually d headed by 
a single individual in whose person is concentrated all the power 
over major decisions. Why is this so? Do the great water works, which 
characterize the core areas of the hydraulic world and which indeed 
require centralized direction, necessitate autocratic leadership? After 
all, controlled (democratic or aristocratic) governments also initiate 
and maintain huge public enterprises. They muster large and disci- 
plined armies and/or fleets; and they operate thus, for substantial 
periods of time, without developing autocratic patterns of rulership. 

Manifestly, the rise of autocratic power depends on more than the 
existence of large state enterprises. In all hydraulic societies proper 
such enterprises play a considerable role; and there, as well as in the 
institutional margin, we always find disciplined armies and almost 
always, also, comprehensive organizations of communication and in- 
telligence. But there is no technical reason why these various enter- 
prises could not be headed by several leading officials. This is indeed 
the case in controlled governments, whose department chiefs are 
carefully separated from, and balanced against, one another. 

However, despotic states lack appropriate mechanics of outside 
control and internal balance. And under such conditions there de- 
velops what may be called a cumulative tendency of unchecked power. 
This tendency could be countered if all major subsections of author- 
ity were more or less equally powerful. It could be countered if the 
chiefs of the public works, of the army, of the intelligence service, 
and of the revenue system were more or less equally strong in terms 
of organizational, communicational, and coercive power. In such a 
case, the absolutist regime might be headed by a balanced oligarchy, 
a "politburo," whose members would actually, and more or less 
equally, participate in the exercise of supreme authority. However, 
the organizational, communicational, and coercive power of the 
major sectors of any government is rarely, if ever, so balanced; and 
under absolutist conditions the holder of the strongest position, ben- 

d. For a few temporary exceptions, like early India, see below, Chap. 8. 

CHAPTER 4, A 107 

efiting from the cumulative tendency of unchecked power, tends to 
expand his authority through alliances, maneuvers, and ruthless 
schemes until, having conquered all other centers of supreme deci- 
sion, he alone prevails. 

The point at which the growth of government functions precludes 
effective outside control differs in different institutional configura- 
tions. But it may safely be said that whenever this critical point is 
passed, the cumulative strength of superior power tends to result in 
a single autocratic center of organization and decision making. 

The crucial importance of this center is not negated by the fact 
that the supreme power-holder may delegate the handling of his af- 
fairs to a top-ranking assistant, a vizier, chancellor, or prime minister. 
Nor is it negated by the fact that he and/or his aide may lean heav- 
ily for advice and speedy action on selected groups of strategically 
placed and carefully tested officials. The governmental apparatus as 
a whole does not cease to be absolutist because the actual center of 
decision making temporarily, and often in a veiled manner, shifts to 
persons or groups below the ruler. 

The sovereign of an agrobureaucratic state may be completely un- 
der the influence of his courtiers or administrators; but such influ- 
ence differs qualitatively from the institutional checks of balanced 
power. In the long run the head of a controlled government must 
adjust to the effective nongovernmental forces of society, while the 
head of an absolutist regime is not similarly restricted. Simple self- 
interest urges any intelligent despot to listen to experienced persons. 
Councillors have existed in most agromanagerial civilizations, and 
not infrequently councils were a standard feature of government. But 
the ruler was under no compulsion to accept their suggestions. 23 

Whether the sovereign was his own chief executive, whether he 
delegated many of his functions to a vizier, or whether he or his 
vizier largely followed the advice of official and nonofficial advisors 
depended, in addition to custom and circumstance, on the personali- 
ties of the ruler and his aides. But despite significant bureaucratic 
attempts to subordinate the absolutist sovereign to the control of his 
officialdom, the ruler could always rule, if he was determined to do 
so. The great monarchs of the Oriental world were almost without 
exception "self-rulers" — autocrats. 

3. Laws of Nature and Patterns of Culture — No 
Effective Checks Either 

Serious observers do not generally contest these facts. However, not 
a few among them seek to minimize their significance by reference 


to mores and beliefs, which are assumed to restrict even the most 
tyrannical regime. 

Mores and beliefs do indeed play a role; and so, for that matter, 
do the laws of nature. However, the potential victims of despotic 
power seem to find little consolation in either fact. They know that 
their masters' behavior, like their own, is affected by the laws of na- 
ture and by more or less firmly established cultural circumstances. 
But they know also that, nevertheless and in the last analysis, their 
fate will be determined by the will of those who wield total power. 

The mechanics of administration and coercion depend on man's 
insight into the laws of nature and his ability to use them. A despotic 
regime will proceed in one way in the neolithic period, in another 
in the iron age, and in still another in our own time. But in each 
case the ruling group asserts its total superiority under the then 
actual natural conditions and by means of the then available tech- 
nology. The victim of a crude form of despotism does not consider 
his persecutors less powerful because, under more advanced technical 
conditions, they may catch and destroy him by different methods 
or with greater speed. 

Nor does he doubt their absolute superiority because they act in 
conformity with prevailing cultural patterns. Such patterns always 
shape the manner in which the ruler (and his subjects) act; and 
occasionally they mitigate or prolong governmental procedures at 
particular stages. But they do not prevent the government from 
ultimately achieving its goal. The fact that in many countries persons 
under sentence of death are normally not executed in certain 
seasons or on certain days 2 * does not mean that they escape their 
doom. And the fact that a dominant religion praises acts of mercy 
does not mean that it refrains from invoking measures of extreme 

The potential victim of despotic persecution knows full well that 
the natural and cultural settings, whatever temporary respites they 
may provide, do not prevent his final destruction. The despotic 
ruler's power over his subjects is no less total because it is limited 
by factors that mold human life in every type of society. 


The power of hydraulic despotism is unchecked ("total"), but it 
does not operate everywhere. The life of most individuals is far 
from being completely controlled by the state; and there are many 
villages and other corporate units that are not totally controlled 

CHAPTER 4, B 10g 

What keeps despotic power from asserting its authority in all 
spheres of life? Modifying a key formula of classical economics, we 
may say that the representatives of the hydraulic regime act (or 
refrain from acting) in response to the law of diminishing adminis- 
trative returns. 

1. The Managerial Variant of the Law of 
Changing Administrative Returns 

The law of diminishing administrative returns is one aspect of what 
may be called the law of changing administrative returns. 1 Varying 
efforts produce varying results not only in a property-based business 
economy ° but also in governmental enterprise. This fact affects 
decisively both the political economy and the range of state control 
in hydraulic society. 

a. Hydraulic Agriculture: the Law of Increasing 
Administrative Returns 

In a landscape characterized by full aridity permanent agriculture 
becomes possible only if and when coordinated human action trans- 
fers a plentiful and accessible water supply from its original loca- 
tion to a potentially fertile soil. When this is done, government-led 
hydraulic enterprise is identical with the creation of agricultural 
life. This first and crucial moment may therefore be designated as 
the "administrative creation point." 

Having access to sufficient arable land and irrigation water, the 
hydraulic pioneer society tends to establish statelike forms of public 
control. Now economic budgeting becomes one-sided and planning 
bold. New projects are undertaken on an increasingly large scale, 
and if necessary without concessions to the commoners. The men 
whom the government mobilized for corvee service may see no 
reason for a further expansion of the hydraulic system; but the 
directing group, confident of further advantage, goes ahead never- 
theless. Intelligently carried out, the new enterprises may involve a 
relatively small additional expense, but they may yield a con- 
spicuously swelling return. Such an encouraging discrepancy ob- 
viously provides a great stimulus for further governmental action. 

b. The Laiu of Balanced Administrative Returns 

The expansion of government-directed hydraulic enterprise usually 
slows down when administrative costs approach administrative 

a. Significantly, the law of diminishing returns has so far been studied primarily in 
connection with private economy (see Clark, 1937: 145 ff.). 


benefits. The upward movement has then reached "Saturation Point 
'A* (Ascent)." Beyond this point further expansion may yield addi- 
tional rewards more or less in proportion to additional administra- 
tive effort; but when the major potentials of water supply, soil, and 
location are exhausted, the curve reaches "Saturation point 'D' 
(Descent)." The zone between Points "A" and "D" is characterized 
by what may be called the law of balanced administrative returns. 

c. The Law of Diminishing Administrative Returns 

Whether Saturation Points "A" and "D" are close together or far 
apart, or whether they coincide, any move beyond this zone of 
balanced returns carries man's action into an area of discouraging 
discrepancy. Here similar, and even increased, administrative en- 
deavors cost more than they yield. It is under these conditions that 
we observe the workings of the law of diminishing administrative 
returns. The downward movement is completed when additional 
outlay yields no additional reward whatsoever. We have then 
reached the absolute administrative frustration point. 

d. Ideal Curve and Reality of Changing Returns 

This ideal curve does not describe the development of any specific 
government-directed system of water works in any specific hydraulic 
society. It indicates in a schematic way the critical points through 
which any hydraulic enterprise passes, if it moves steadily through 
all zones of growing and shrinking returns. 

Rarely, if ever, do the actual and the ideal curves coincide. 
Geology, meteorology, potamology, and historical circumstance make 
for countless variations. Progress toward saturation and beyond may 
be interrupted by longer or shorter countermovements. But every 
section of the curve reflects a genuine trend; and the entire curve 
combines these trends to indicate all possible major phases of 
creation and frustration in hydraulic enterprise. 

e. Nonhydraulic Spheres of Political Economy 

In the sphere of agricultural production itself, coordinated and 
government-directed action yields increasing administrative returns 
only under primitive and special conditions. It is only in techno- 
logically crude hydraulic societies that mass labor on "public" fields 
prevails. And even in these societies the government does not try 
to assume managerial direction over the fields which have been set 
aside for the support of the individual farmer. In a technically more 

CHAPTER 4, B 111 

advanced setting, the administrative creation point and the adminis- 
trative frustration point tend to coincide. For there the hydraulic 
regime prefers to refrain altogether from agricultural production, 
which from the standpoint of administrative returns is more reason- 
ably handled by many small individual farming units. 

Of course, political needs take precedence over economic consider- 
ations. The great agromanagerial enterprises of communication and 
defense are cases in point, as are certain government-run workshops 
(arsenals, shipyards). However, the hydraulic regime's reluctance to 
assume direct control over the finishing industries derives from the 
realization that in this field state management would involve deficits 
rather than gains. In hydraulic as well as in other agrarian societies 
the government is therefore satisfied to leave the bulk of all handi- 
craft to small individual producers. 

2. The Power Variant of the Law of Changing 
Administrative Returns 

a. Imperative and Worth-while Efforts 

It is easy to recognize the workings of the law of changing adminis- 
trative returns also in the sphere of political power. The efforts of the 
hydraulic regime to maintain uncontested military and police con- 
trol over the population prove increasingly rewarding until all 
independent centers of coercion are destroyed. The expenses incurred 
in supporting speedy communications and intelligence follow a 
similar pattern; and the expansion of fiscal and judicial action ap- 
pears reasonable as long as it satisfies the rulers' desire for uncon- 
tested political and social hegemony. 

Some of these operations are imperative, others at least worth 
while. But carried beyond Saturation Point "D", they all become 
problematic. The discouraging discrepancy between continued en- 
deavor and decreasing political rewards makes the government 
reluctant to use its apparatus much below this point. 

b. The Forbidding Cost of Total Social Control in a 
Semimanagerial Society 

The developed industrial apparatus state of the USSR has crushed 
all independent nationwide organizations (military, political, pro- 
prietary, religious); and its total managerial economy permits the 
establishment of innumerable bureaucratic bases for controlling all 
secondary (local) professional groupings and even the thought and 
behavior of individuals. The hydraulic apparatus state does not 


have equal facilities. It is strong enough to prevent the growth of 
effective primary organizations; and in doing so, it brings about that 
one-sided concentration of power which distinguishes it from the 
ancient and medieval agrarian societies of the West. But being only 
semimanagerial, it lacks the ubiquitous bases which enable the men 
of the apparatus to extend their total control over secondary organi- 
zations and individual subjects. In the USSR such total control was 
initiated through the nationalization of agriculture (the "collectiviza- 
tion" of the villages); and it was accomplished through the pulveri- 
zation of all nongovernmental human relations. Hydraulic society 
never made the first step, and it therefore never laid the foundations 
for the second. 

To be sure, the notion of a ubiquitous control also attracted the 
master minds of hydraulic despotism. Garcilaso de la Vega, a scion 
of native royalty, claimed that under Inca rule special officials went 
from house to house to make sure that everybody was kept busy. 
Idlers were punished by blows on the arms and legs "and other 
penalties prescribed by the law." 2 The great Chinese "Utopia" of 
bureaucratic government, the Chou Li, lists several officials who, 
in a well-managed state, should regulate the people's life in village 
and town. 

There is no reason to doubt that the Incas wanted their subjects 
to work as much as possible; but any effective inspection of the 
commoners' domestic life would have required an army of officials, 
which would have eaten up a great part of the public revenue with- 
out providing a compensatory increase in income. It is therefore hard 
to believe that the "laws" mentioned by Garcilaso went far beyond 
a general — and therefore not too costly — supervision. The same may 
be said for the classic book of Chinese bureaucracy. All educated 
Chinese officials studied the Chou Li; but once in office, they soon 
learned to distinguish between the sweet dream of total social con- 
trol and the sober administrative reality. Except for some short-lived 
attempts at extreme interference, they were content to maintain firm 
control over the strategically important spheres of their society. 

c. Total Social Control Not Necessary for the Perpetuation 
of Agromanagerial Despotism 

To say that the law of diminishing administrative returns dis- 
courages the hydraulic state from attempting to control individuals 
and secondary organizations totally is only another way of saying 
that the government feels no fundamental need to do so. If it were 
otherwise — that is, if total control were imperative for the perpetua- 

CHAPTER 4, B 113 

tion of the despotic regime — the rulers might have to spend all 
their income to be safe. Obviously, such a power system would be 

Historical experience shows that during long periods of "peace and 
order" the hydraulic rulers can maintain themselves without resort- 
ing to excessively costly measures. It also shows that under "normal" 
conditions they need not make severe material sacrifices. Except in 
times of unrest, they are adequately protected by their wide-flung 
network of intelligence and coercion, which successfully blocks the 
rise of independent nationwide primary organizations and prevents 
discontented individuals or secondary organizations from gaining 

The political crises that develop periodically may be caused in 
part by the dissatisfaction of such individuals and organizations. 8 
But serious unrest, whatever its origin, soon assumes a military form, 
and it is combated by outright military measures. Responding to 
the law of diminishing administrative returns, the masters of the 
agrarian apparatus state run the risk of occasional uprisings and do 
what their modern industrial successors do not have to do: they 
grant a certain amount of freedom to most individuals and to certain 
secondary organizations. 

3. Sectors of Individual Freedom in 
Hydraulic Society 

a. Limitations of Managerial Control 

The duration of the state corvee determines the period during 
which a member of hydraulic society is deprived of his freedom of 
action. The corvee may have many objectives, but it must allow 
the mass of the laborers — the peasants — sufficient time to attend 
to their own economic affairs. Of course, even in the villages the 
peasants may have to submit to a policy of economic planning, but 
at most this policy involves only a few major tasks, such as plowing, 
sowing, harvesting, and perhaps the choice of the main crop. Often 
it does not go this far; and at times it may be altogether absent. 

Under conditions of advanced technology the corvee also tends 
to change and shrink. Work on the public fields may be replaced 
by a tax; and larger or smaller segments of the nonagricultural corvee 
may be similarly commuted. 

But whatever the character of the rural communities and whatever 
the duration of the public labor service may be, there are definite 
and at times considerable periods in the peasant's life during which 


he proceeds at his own discretion. This is still more true for the 
nonagrarian commoners. Artisans and traders who, in a differentiated 
societal setting, pursue their occupations professionally and pri- 
vately * may become more valuable as taxpayers than as corvee 
laborers. Their freedom of movement will increase correspondingly. 
Marx speaks of the "general slavery" of the Orient. According to 
him, this type of slavery, which is inherent in man's attachment 
to the hydraulic commonwealth and state, 5 differs essentially from 
Western slavery and serfdom. 6 The merit of Marx' formula lies in 
the problem it raises rather than in the answer it gives. A person 
commandeered to toil for an "Asiatic" state is a slave of the state as 
long as he is so occupied. He is perfectly aware of the lack of free- 
dom, which this condition involves, and he is equally aware of the 
pleasure of working for himself. Compared with the total state 
slavery of the total managerial industrial society, the partial state 
slavery of the partial managerial hydraulic society makes indeed 
considerable concessions to human freedom. 

b. Limitations of Thought Control 

A comparable tendency to make concessions arises also in the 
sphere of thought control. To appreciate fully what this means, we 
must understand the enormous stress that the masters of the hydrau- 
lic state place on the society's dominant ideas. The close coordina- 
tion of secular and religious authority makes it easy to apply this 
stress to both the higher and the lower strata of society. The sons 
of the dominant elite are generally educated by representatives of 
the dominant creed; and the whole population is in continued and 
government-promoted contact with the state-attached temples and 
their priesthoods. 

Education usually is a long process, and its influence is profound. 
In India the young Brahmin who prepared himself for priestly 
office had to study one, two, or all three Vedas, applying himself to 
each one of them for twelve long years. And the members of the 
"protecting" Kshatriya caste, and even those of the next lower caste, 
the Vaisya, were also advised to study the Sacred Books. 6 In China 
"learning" — the study of the canonical (classical) writings — was al- 
ready considered a basic prerequisite for administrative office in 
Confucius' time. 7 Increasing systematization led to the holding of 

b. Marx assumed that from the European point of view, in this general Asiatic 
slavery, the laborer seems to be a natural condition of production for a third person 
or a community, as under [private-property-based] slavery and serfdom, but that 
actually "this is not the case" (Marx, 1939: 395). 

CHAPTER 4, B 115 

elaborate and graded examinations, which fostered perpetual ideo- 
logical alertness in all energetic and ambitious young, and in many 
middle-aged and even elderly, members of the ruling class. 

But the same societal forces that led to the systematic perpetua- 
tion of the dominant ideas also encouraged a variety of secondary 
religions. Many simple hydraulic civilizations tolerated independ- 
ent diviners and sorcerers, 8 whose artisan-like small-scale activities 
modestly supplemented the coordinated operations of the leading 
tribal or national creed. Under more complex conditions, ideo- 
logical divergence tended to increase. Often the subject of a hydraulic 
state might adhere to a secondary religion without endangering his 
life. Non-Brahministic creeds, such as Jainism or Buddhism, are 
documented for India from the first millennium b.c. Buddhism per- 
sisted in traditional China, despite temporary persecutions, for al- 
most two thousand years. And the Islamic Near East, India, and 
Central Asia were similarly indulgent. 

In the ideological as in the managerial sphere, the policies of the 
agrarian apparatus state contrast strikingly with policies of the 
modern industrial apparatus states, which, while feigning respect 
for traditional ("national") culture and religion, spread the Marxist- 
Leninist doctrine with the avowed aim of eventually annihilating 
all other ideologies. Again, the difference between their policies is 
not due to any innate tolerance on the part of the agrobureaucratic 
rulers, whose insistence on the unique position of the dominant 
religion is always uncompromising and frequently ruthless. But the 
law of diminishing administrative returns places an exorbitant price 
on the attempt to maintain total ideological control in a differ- 
entiated semimanagerial society. And here, as in the operational 
sector, experience shows that the absolutist regime can perpetuate 
itself without making so costly an effort. 

4. Groups Enjoying Varying Degrees of Autonomy 

Experience shows still more. It assures the hydraulic rulers that 
they may — for the same reasons — permit some autonomy not only 
to their individual subjects but to certain secondary groups as well. 
In referring to heterodox creeds, we are aware that their adherents 
are usually permitted to establish congregations, which support 
either individual priests or larger or smaller priesthoods. Since the 
early days of written history, the artisans and traders of hydraulic 
civilizations have formed professional organizations (guilds). More 
ancient still are the village communities, which have probably existed 
as long as hydraulic civilization itself. Kin groups are institutionally 


older than agriculture; and like the village community, they are 
present everywhere in the hydraulic world. 

These types of associations differ greatly in distribution, composi- 
tion, quality, and purpose. But they have one thing in common. All 
of them are tolerated by the despotic regime. Many supervisory 
measures notwithstanding, they are not subjected to total control. 

a. Less Independence than Frequently Assumed 

Romantic observers have taken the absence of such control as 
evidence for the existence of genuine democratic institutions in the 
lower echelons of hydraulic society. In this form, the claim cannot 
be accepted. Throughout the hydraulic world, government authority 
and family authority are interlinked; and measures of political con- 
trol affect the majority of all villages, guilds, and secondary religious 

Parallels can be found in other agrarian societies for most of these 
restrictive trends. (The free guilds of feudal Europe are as ex- 
ceptional as they are significant.) This, however, is not the issue 
here. What we are concerned with is whether, in contrast to corre- 
sponding developments in other despotic states — and also in contrast 
to restrictive developments in other agrarian civilizations — the sec- 
ondary organizations of hydraulic society were genuinely autono- 
mous. The answer to the question is "No." 


The family of traditional China has often been said to be the 
institution that gave Chinese society its peculiar character and 
strength. This thesis is correct insofar as it stresses the family as a 
basic component of society; but it is misleading insofar as it implies 
that the family determined the quality and power of the institutional 
setting of which it was a part. 

The authority of the Chinese pater familias was much stronger 
than intrafamilial leadership required; e and he owed his extraor- 
dinary power essentially to the backing of the despotic state. Dis- 
obedience to his orders was punished by the government. 9 On the 
other hand, the local officials could have him beaten and imprisoned, 
if he was unable to keep the members of his family from violating 
the law. 10 Acting as a liturgical (semi-official) policeman of his kin 
group, he can scarcely be considered the autonomous leader of an 
autonomous unit. 

c. For the nongovernmental roots of paternal authority in the Chinese family see 
Wittfogel, 1935: 49; ibid., 1936: 506 ff. 

CHAPTER 4, B liy 

The Babylonian father, who could place his wife, son, or daughter 
in the service of a third person for several years, 11 also owed his 
power to the government which backed him up in his decision. 
Whether he was legally responsible for the behavior of the family 
members is not clear. 

The patria potestas of ancient Egypt has been compared with that 
of Rome. The strongly militarized society of republican Rome did 
indeed encourage the development of highly authoritarian family 
relations; but the Egyptian father seems to have had still greater 
power than his Roman counterpart.* 

In the Islamic world, respect for the parents is prescribed by the 
Sacred Law; 12 and the degree to which paternal authority operated, 
particularly in the villages, may be judged from the fact that in 
such countries as Syria the father customarily was the master over his 
family until his death. 13 

The Law Books of India give the father an almost kinglike power 
over members of his kin group. 14 Despite several restrictions, 16 his 
authority over his wife and children seems to have been extremely 

Evidently the father's power varied notably in different hydraulic 
civilizations. But almost everywhere the government was inclined 
to raise it above the level suggested by his leadership functions in 
the family. 


Generally the villages of hydraulic civilizations are under the 
jurisdiction of headmen who are either government-appointed or 
elected by their fellow villagers. Appointment seems to be frequent 
in the regulated rural communities of compactly hydraulic civiliza- 
tions, whereas free choice is more apt to be permitted in less 
compactly hydraulic societies. In Inca Peru the local officials down 
to the lowest functionary — the head of ten families — was appointed. 18 
In pre-Conquest Mexico, too, the village land was communally 
regulated. But its agrarian economy was much less bureaucratized 

d. Dr. Taubenschlag's assertion that the Egyptian father's right to sell his child has a 
Roman counterpart is documented only for "the fourth century" (Taubenschlag, 1944: 
103 ff.). 

e. Jolly, 1896: 78. At the beginning of the 19th century, Dubois (1943: 307 ff.) found 
the authority of the Brahmins enormous, whereas paternal authority was weak. The 
author lived in India from 1792 to 1823. Assuming that he observed the phenomenon 
correctly, we are at a loss to explain it. Was it, at least in part, due to the turmoil 
of the time? 


than that of the Inca empire. The heads of the Mexican local ad- 
ministrative units, the calpulli, were elected. 17 

However, this correlation does not prevail generally, perhaps be- 
cause appointment is only one among several ways of controlling a 
local functionary. Almost everywhere the hydraulic government 
holds the headman responsible for the obligations of his co-villagers. 
It thus places him in a position of state dependency. Where land is 
communally held and where taxes are communally paid, the village 
headman is likely to wield considerable power. Assisted by a scribe 
and one or several policemen, he may become something of a local 

The inscriptions of the early Near East show the regional officials 
actively concerned with plowing and the collection of the reve- 
nue; 1S but we are unable to get a clear picture of how the village 
functionaries fitted into the administrative nexus. 19 As in other 
spheres of life, the Persians and their Hellenistic and Roman succes- 
sors may well have perpetuated an earlier village pattern. In 
Ptolemaic and Roman Egypt the leading village official, the scribe, 
assisted by the elders, executed his government-imposed tasks. 20 
These men, no matter whether they were appointed 21 or elected 
like the elders, 22 were all "directly dependent on the central govern- 
ment . . . they all especially obeyed the strategos of the district." 23 

The data for Roman Syria seem to suggest considerable popular 
participation in village affairs, 24 whereas the Egyptian village officials 
probably acted in a very authoritarian manner. But this divergence 
must not make us overlook the basic similarities that existed through- 
out the ancient Near East in village organization and government 
dependency. 25 In Hellenistic times, 26 as previously, the "royal" 
villagers were attached to the land they cultivated. 27 It therefore 
seems safe to conclude that in the pre-Roman as well as in the Roman 
period the peasants of Syria and Asia Minor did not administer their 
villages autonomously. 

In Arab Egypt, as in Byzantine Egypt, 28 the village administration 
was in the hands of a headman and the elders. Under the Arabs 
the headman, who possibly was nominated by the peasants and con- 
firmed by the government, 29 seems to have apportioned and collected 
the tax. 30 He designated the corvee laborers and exercised police and 
judicial functions. 81 

In the Arab provinces of the Turkish Near East the village head- 
man (sheikh) assisted the official and semi-official representatives of 
the government in allocating the tax. 32 He "policed the fellahs who 
cultivated the lands under his charge, and the principal seyh acted 
as magistrate and arbitrator, with authority not only over the culti- 

CHAPTER 4, B lig 

vators but over all the inhabitants." 3S Controlling "his" peasants in 
an arbitrary way and being in turn controlled with equal severity by 
the state bureaucracy, 84 he certainly was not the representative of a 
free rural village community. 

In India the village headman may have been elected originally; s5 
but from the time of the later Law Books on — that is, from the end 
of the first millennium B.C. — his appointment is documented. 36 As 
the king's representative in the villages, who "collected taxes for 
him" 87 and who also fulfilled policing and judicial functions, 88 the 
headman held a position of authority not dissimilar to that enjoyed 
by his Near Eastern counterpart. Muslim rule did not fundamentally 
change this administratively convenient arrangement, which in fact 
persisted in the majority of all Indian villages up to modern times. 89 

In China the regulated village yielded to a property-based pattern 
more than two thousand years ago. The duties of the village officials 
shrank correspondingly, but they did not disappear altogether. At 
the close of the imperial period most sizable villages had at least two 
functionaries, a headman, chuang chang, and a local constable, t i fang 
or ti pao. 40 The headman, who was usually chosen by the villagers, 
executed the directing, and the constable, who usually was govern- 
ment appointed/ the coercive, functions of the village government. 
They cooperated in their official tasks: the collection of taxes and 
materials for public constructions, the organizing and directing of 
corvee services ("government transportation . . . work on river- 
banks, patrols for the Imperial roads" etc.), 41 and the making of 
intelligence reports. 42 

All these activities linked the headman to the central government, 
although he was not part of its bureaucracy." The villagers found it 
hard to bring a complaint against him, even if their case was good, 
for he monopolized communication with the district magistracy. 43 
The constable was controlled by the county officials. They could 

/. According to Smith (1899: 227). the candidates for this position were "not formally 
chosen, nor formally deposed." Instead they used to "drop into their places" as the 
result of what Smith calls "a kind of natural selection." 

It would probably be better to speak of an informal election based on an under- 
standing between all family heads of some standing. Dr. K. C. Hsiao, who has almost 
completed his comprehensive study, Rural China, Imperial Control in the Nineteenth 
Century, ascribes "a certain amount of informal local influence on village leader- 
ship," especially that of "wealthy or gentry families." But he finds it impossible to 
give quantitative data about "the proportion of government-appointed village head- 
men (pao-chang, chia-chang, etc.; and later, chuang-chang, ti-pao, ti-fang, etc.)." He 
adds: "The official scheme called for universal institution of such headmen, wherever 
rural communities existed" (letter of January 15, 1954). 

g. Usually the village paid him a salary (Werner, 1910: 106 ff.). In addition there 
were the usual material advantages inherent in the handling of public money. 


have him "beaten to a jelly" for neglecting his duty as a local in- 
telligence agent. 44 

The villages of imperial China were less strictly controlled than 
those of pre-Conquest Peru, India, and most Near Eastern civiliza- 
tions, but even they did not govern themselves. Their main func- 
tionaries, who were either appointed or confirmed by the govern- 
ment, were inescapably tied to an operational system that served 
the interests of the government rather than the interests of the 


The professional corporations of the artisans and traders in hydrau- 
lic civilizations were similarly conditioned. Again the appointment 
of the leading official is significant; but again it is only one of several 
ways in which the despotic state assures its unchecked superiority 
and the weakness of the tolerated organization. 

Hellenistic Egypt seems to have followed ancient usage in having 
persons "working for the State in industry, transport, mining, build- 
ing, hunting, etc." gathered into professional groups that were 
"organized and closely supervised by the economic and financial ad- 
ministration of the king." 45 

In the later part of the Roman empire and in Byzantium, the 
government "strictly regulated" the activities of the guilds.* Until 
the third century the members elected their own headmen; but from 
that time on the government made the final decision on guild- 
nominated headmen, who, after installation, were supervised and 
disciplined by the state. 46 

In Ottoman Turkey officials inspected the markets 47 and con- 
trolled the prices, weights, and measurements,* thus fulfilling func- 
tions which in the burgher-controlled towns of Medieval Europe 
were usually the responsibility of the urban authorities. 48 Further- 
more, the state, which in most countries of feudal Europe collected 
few if any regular taxes from the urban centers of strongly developed 
guild power, was able in Turkey to tax the guilds and, as elsewhere 
in the Orient, to employ as its fiscal agents the headmen of these 
corporations, who "distributed the tax-quotas of their members" and 
who were "personally responsible for their payment." 49 

In Hindu India, the setthi, the head of the merchant guild, was 
a semi-official closely attached to the ruler's fiscal administration. 50 

h. Stockle, 1911: 11. For reference to guild heads as tax collectors in Byzantine and 
Arab Egypt, see Grohmann, PAP: 279 and n. 8. For conditions at the beginning of 
Arab rule, see ibid.: 131, n. 3, and Crum, 1925: 103-11. 

i. Specifically this was done by agents of the kadi (Gibb and Bowen, 1950: 287). 

CHAPTER 4, B 121 

The merchants represented considerable wealth, and their corpora- 
tions seem to have been more highly respected than those of the 
artisans. 51 But this did not make the merchant guild a significant 
political entity. 

It has been said that the Indian guilds came into prominence 
in early Buddhist days. 52 In agreeing with this observation, however, 
we must be careful not to exaggerate its political significance. Accord- 
ing to Fick, "the corporations of the manufacturers fall — partly at 
any rate — undoubtedly under the category of the despised castes"; 53 
and Dr. Rhys-Davids insists that there is "no instance as yet produced 
from early Buddhist documents pointing to any corporate organisa- 
tion of the nature of a gild or Hansa league." 5i A legend of the 3d 
or 4th century, which is supposed to show that the town of Thana ' 
was "ruled by a strong merchant guild" actually describes the un- 
successful attempt of a group of merchants to combat a competitor 
by cornering the market.* 

In China the existence of guilds is reliably documented only since 
the second half of the first millennium a.d. Under the T'ang and 
Sung dynasties the guild heads could be held responsible for the im- 
proper professional behavior of their members, such as violations of 
the currency regulations, 55 theft, and other misdeeds. And in many 
cases membership was compulsory. 56 The guilds as a unit also had 
to render special services to the state. 57 In recent centuries the gov- 
ernment seems to have left the less significant craft and trade guilds 
largely to their own devices; m but the corporations of such important 
groups as the salt merchants n and a number of Cantonese firms 
dealing in foreign trade ° were strictly supervised. 


Our information on secondary religions is particularly plentiful for 
Islamic society and traditional China. Muslim rulers tolerated Chris- 
tianity, Judaism, and Zoroastrianism." But followers of these creeds 
had to accept an inferior status both politically and socially, and 

/. Poona, south of modern Bombay. 

k. Hopkins, 1902: 175. Hopkins' erroneous thesis is taken up by Max Weber in an 
argument stressing the temporary political prominence of the Hindu guilds (Weber, 
RS, II: 86 ff.). See below, p. 266. 

m. Wittfogel, 1931: 580 ft., 714 ff. My 1931 analysis overlooked the state-controlled 
guilds of important trades, such as the salt business. 

n. The guild heads collect the tax from the "small merchants" (Ch'ing Shih Kao 
129. ib). 

o. The headmen were appointed by the government (Yueh Hai Kuan Chih 25. 2a). 

p. Macdonald, 1941: 96; Grunebaum, 1946: 117. Zoroastrians were tolerated originally 
(Mez, 1922: 30); later they were more harshly treated (Buchner, 1941: 381). 


they were prevented from spreading their ideas. The laws forbade 
conversion from Christianity to Judaism or vice versa; and penalties 
for apostasy from Islam were severe. Christians were not permitted 
to beat their wooden boards loudly, 9 or sing in their churches with 
raised voices, or assemble in the presence of Muslims, or display 
their "idolatry," "nor invite to it, nor show a cross" on their 
churches. 58 No wonder that the religious minorities — who during 
the Turkish period were set apart in organizations called millet 59 — 
vegetated rather than throve. The head of the millet was nominated 
by the millet r but appointed by the sultan; 60 once in office he 
was given "just enough executive power ... to enable him to 
collect the taxes imposed on his community by the state." 81 

In traditional China, Buddhism was the most important secondary 
religion. It reached its greatest prominence in the barbarian dynasties 
of infiltration and conquest which ruled over the old northern 
centers of Chinese culture during the middle period of the first 
millennium a.d. 62 The harsh persecutions of 845 initiated a policy 
which over time reduced it to a carefully restricted secondary 

Specially designated officials supervised Buddhism and other prob- 
lematic creeds. 83 The government limited the erection of monasteries 
and temples; 64 it licensed the number of priests and monks; 65 it 
forbade certain religious activities which in other countries went 
unrestricted; and it prescribed that "the Buddhist and Taoist clergy 
shall not hold sutra-readings in market-squares, nor go about with 
alms-bowls, nor explain the fruits of salvation, nor collect moneys." 6e 
Concluding his classical survey of what others have hailed as the 
elements of religious liberty, Be Groot asks: "What is the good of 
this liberty where the State has cast its system of certification of 
clergy within such strict bounds, and has made the admission of 
male disciples extremely difficult, of females almost impossible, so 
that the number of those who could avail themselves of such liberty, 
is reduced to a miserably small percentage of the population? It 
makes this vaunted liberty into a farce." 67 

b. Genuine Elements of Freedom Nevertheless Present 

Thus the hydraulic state restrictively affects practically all secondary 
groups and organizations, but it does not integrate them completely 
into its power system. 

The traditional Chinese family, whose head enjoyed a particularly 

q. These boards were used as bells (Grunebaum, 1946: 179). 
r. Or its clergy? 

CHAPTER 4, B 123 

distinguished position legally, was not forced by political and police 
pressure to set one family member against another, as is the case in 
modern apparatus states. In China and in India the government 
permitted the kin groups to settle their internal affairs in accordance 
with their own family "laws." f8 In other hydraulic civilizations the 
families enjoyed a less formal, but equally effective, quasi-autonomy. 

Government control over the villages, although very specific, is 
also definitely limited. Even where village officials wield much 
power, the peasants who live alongside them have many opportunities 
to make their opinions on the day-to-day affairs of the community 
felt. And once the demands of the government are satisfied, the 
headman and his aides usually settle the affairs of their village with 
little, if any, interference from above. 

Certain opportunities for self-government seem to have existed in 
the villages of Roman Syria 69 and in the Egyptian villages of the 
Roman and Byzantine period. 70 The village chief of Ottoman 
Turkey, like his counterparts in other Oriental civilizations, acted 
with great independence as far as the internal affairs of the rural 
community were concerned. 71 

The headman of an Indian village could fulfill his functions suc- 
cessfully only by trying "to conciliate the villagers." 72 He could not 
be "proud, intolerant, and haughty like the Brahmins"; instead he 
had to be "polite and complaisant" toward his equals and "affable 
and condescending" toward his inferiors. 73 Full-fledged committee 
organizations were probably confined to the small minority of rural 
settlements dominated by landholding groups, primarily Brahmins. 74 
But the informal assembly (panchayat) of village elders or all 
villagers is said to have been a general institution; 75 and its meet- 
ings apparently softened the authority of the headman. Since the 
villages, except for official demands, remained more or less in the 
charge of the headmen and their aides, they were indeed rural is- 
lands, enjoying partial autonomy. 76 

In the traditional Chinese village the local officials were still 
closer to the nonofficiating co-villagers, who, particularly when they 
belonged to wealthy or gentry families, might exert great influence 
in local affairs. 77 Criticism from an "out" group of fellow villagers 
might compel the headman and his supporters to resign. Under such 
pressure, a "band of men" who had been in power for a long time 
might withdraw "from their places, leaving them to those who 
offered the criticisms." 78 

Such behavior does not imply a formal democratic pattern; but 
it has a democratic flavor. Of course, there are various kinds of 
official requests; and there is always the constable, and often a tax 


collector, both government appointed and both spectacularly repre- 
senting the interests of the bureaucratic apparatus. But here outside 
control usually ends. The government "places no practical restric- 
tions upon the right of free assemblage by the people for the con- 
sideration of their own affairs. The people of any village can if 
they choose meet every day in the year. There is no government 
censor present, and no restriction upon liberty of debate. The 
people can say what they like, and the local Magistrate neither knows 
nor cares what is said." T9 

In many hydraulic civilizations the government was as little 
concerned about the internal affairs of the guilds. The Indian Law 
Books advised the king to recognize the statutes (laws) of the 
guilds. 80 And similar statutes existed elsewhere. 81 The Turkish guilds 
were subject to "the overriding authority of the temporal and 
spiritual powers, represented by governors, police officers, and 
kadis"; 82 and their headmen were held responsible by the govern- 
ment for the execution of its fiscal tasks. However, otherwise and 
"within the limits imposed by religion, tradition, and 'usage,' . . . 
the corporations were relatively free and autonomous." 83 Gibb and 
Bowen therefore list them among "the almost self-governing 
groups." 84 

Gibb's and Bowen's formula is valid also for the secondary 
religions. All external restrictions notwithstanding, these religions 
did enjoy "some fragments of religious liberty." In traditional China 
the priests of the secondary religions, "seeking their own and other 
people's salvation, are not forbidden to preach, recite sutras, and 
perform ceremonies within doors." 85 And under Islam, "each non- 
Moslem congregation administers its own affairs under its responsible 
head, a rabbi, bishop, etc." 86 As long as their worship disturbed no 
"true believers," and as long as their organization presented no 
security threat, the government usually permitted the religious 
minorities to live, within their congregations, a more or less autono- 
mous life. 

5. Conclusion 

a. Politically Irrelevant Freedoms 

These are indeed modest freedoms! They occur in varying com- 
binations in several spheres of life. And by now we should be able 
to understand why they do occur, and why they are so limited. 

Hydraulic society is certainly not immune to rebellious move- 
ments, but kin organizations even in their extended forms are no 
political threat to a normally functioning agrobureaucratic des- 

CHAPTER 4, B 125 

potism. Nor are the villages a serious threat. The relatively far- 
reaching autonomy of the traditional Chinese village could, in case 
of an insurrection, "be extinguished in a moment, a fact of which 
all the people are perfectly well aware." 8T Secondary religious groups 
might be a danger in times of great unrest. And this is probably why 
the government of imperial China never relaxed its control over 
the tolerated creeds and was so ready to suppress certain sects. 88 
The rebellious potential inherent in the guilds was perhaps never 
completely eliminated, but the hydraulic government was able to 
paralyze it without exhausting its revenues. 

Grunebaum finds it "remarkable to observe how little the Muslim 
state was really hampered in its operation by the dead weight of 
these semi-foreign organizations within its structure." 89 And others 
have commented in the same vein on the political effect of guilds in 
hydraulic civilizations. The early Byzantine state had no need to 
liquidate the still-existing Roman guilds, "since they were not at 
all dangerous politically, and since they could exert no pressure 
whatsoever on the government and administration, as did, for in- 
stance, the German guilds of the Middle Ages." 90 Massignon, who 
more than most of his colleagues considers the Muslim guilds at 
least temporarily a political factor, is nevertheless aware that they 
"never attained a political influence comparable to that of the 
medieval European guilds." 91 Gibb and Bowen consider the powers 
of the medieval guilds in Europe so much broader than those of 
the Islamic corporations that they doubt the suitability of the very 
term "guild" for the latter. 92 An equation between the guilds of the 
Medieval West and the guilds of India 9S or of China 94 has been 
rejected for similar reasons. 

To be sure, there existed many resemblances between the two types 
of corporations, resemblances created by the peculiarities and needs 
of the organized professions; 9S but the profoundly different societal 
settings in which they operated gave them profoundly different po- 
litical and social qualities. The guildsmen of the later European 
Middle Ages frequently became the masters of their towns; and as 
such they might play an active part in the power struggles of their 
time. The guildsmen of the hydraulic world were permitted a cer- 
tain autonomy, not because, politically speaking, they were so strong, 
but because they were so irrelevant. 

b. A Beggars' Democracy 

I n modern totalitarian states the inmates of concentration and forced 
labor camps are permitted at times to gather in groups and talk at 
will; and not infrequently certain among them are given minor 


supervisory jobs. In terms of the law of diminishing administrative 
returns such "freedoms" pay well. While saving personnel, they in no 
way threaten the power of the commandant and his guards. 

The villages, guilds, and secondary religious organizations of agro- 
managerial society were no terror camps. But like them they enjoyed 
certain politically irrelevant freedoms. These freedoms — which in 
some instances were considerable — did not result in full autonomy. 
At best they established a kind of Beggars' Democracy. 


1. Total Power — for the Benefit of the People? 

The hydraulic state is not checked by a Beggars' Democracy. Nor is 
it checked by any other effective constitutional, societal, or cultural 
counterweights. Clearly it is despotic. But does it not at the same time 
benefit the people? 

2. The Claim and the Reality 

a. Operational Necessity Not to Be Confused with Benevolence 

The hydraulic state is a managerial state, and certain of its opera- 
tions do indeed benefit the people. But since the rulers depend on 
these operations for their own maintenance and prosperity, their 
policies can hardly be considered benevolent. A pirate does not act 
benevolently when he keeps his ship afloat or feeds the slaves he 
plans to sell. Capable of recognizing his future as well as his present 
advantages, he is rational but not benevolent. His behavior may 
temporarily benefit the persons in his power; but this is not its pri- 
mary purpose. Given a choice, he will further his own interests, and 
not the interests of others. 

b. The Rationality Coefficient of Hydraulic Society 

On the level of total power, the representatives of hydraulic regimes 
proceed in a similar way. Their behavior may to some degree benefit 
the persons in their power, and far-sighted advisors and statesmen 
may stress the importance of satisfying the people; "but taken as a 
group they consider the needs of their subjects in the light of their 
own needs and advantages. For this purpose they must (i) keep the 
agrarian economy going; (2) not increase corvee labor and taxes to a 

a. For India see Bhagavadgita, passim, and Manu, 1886: 229, 396 ff. For China: the 
sayings of Confucius and still more important, those of Mencius. 

CHAPTER 4, C 127 

point where the discouraged peasants stop producing; and (3) not 
permit internal and external strife to disrupt the life of the popula- 

The third task — the maintenance of peace and order— confronts 
the governments of all societies. The first and second tasks distin- 
guish hydraulic from other agrarian civilizations. The continued ex- 
istence of agrarian despotism depends on the satisfactory execution 
of these three functions. They constitute what may be called the 
regime's rationality minimum. 

Conquest societies, whose rulers are steeped in nonhydraulic tra- 
ditions, often proceed along or near the lowest hydraulic rationality 
level. And endogenous masters frequently sink to this level during 
periods of decay and disintegration. Strong moves toward a higher 
rationality coefficient occur particularly during the earlier phases of 
endogenous rule, but they may also occur during later periods of 
growth or consolidation. 

The formative phase of a conquest society is largely determined by 
the conquerors' ability to identify themselves with their new institu- 
tional environment. The Mongols were completely alien to the tradi- 
tions and mores of the hydraulic civilizations they overran. Chingis 
Khan's son, Ogotai, is said to have planned to convert the cultivated 
fields of China into pastures; and he refrained from doing so only 
because Yeh-lii Ch'u-ts'ai convincingly explained to him the superior 
tax potential of the agrarian order. x But although the Mongols main- 
tained the hydraulic economy of their new realm, they remained in- 
different to its subtler needs. Virtually everywhere they stayed close 
to the rationality minimum of hydraulic society. 

Mohammed, who lived in arid Arabia, certainly understood the 
importance of irrigation for successful crop-raising, although in his 
official utterances he rarely refers to the problem, and then essen- 
tially to small-scale (well) irrigation. 2 His followers preserved, re- 
stored, and even created vigorous hydraulic economies in Syria, 
Egypt, Iraq, Northwest Africa, Spain, and briefly also in Sicily. The 
Manchus were familiar with irrigation agriculture before they moved 
southward across the Great Wall to conquer China. 3 In this respect 
they were not unlike the Incas, who practiced irrigation in the An- 
dean highlands before they established their hydraulic empire. 4 
When they were overrun by the Spaniards, they were probably op- 
erating close to their rationality maximum. 

c. Whose Rationality Coefficient? 

But no matter whether a hydraulic society is operated crudely or 
subtly, the claim of benevolence compels us to ask: cui bono? Evi- 


dently operational tasks may be handled in a way that satisfies the 
interests of the rulers at the expense of the nongovernmental forces 
of society. Or they may be handled in a way that satisfies the needs 
of the people and gives few, if any, advantages to the government. 
Intermediate solutions compromise between the two extremes. 

As a rule, the three alternatives are seriously considered only if 
the actual circumstances permit genuine choice. In the managerial, 
the consumptive, and the judicial spheres of hydraulic life this is 
indeed the case. But in all these spheres we find the people's interests 
sacrificed to the rulers' rationality optimum. 

3. The Rulers' Rationality Optimum Prevails 

a. Necessity and Choice in the Policy of the Hydraulic Regime 

In the territorial states of ancient China, as in other hydraulic civ- 
ilizations, philosophers discussed the alternatives of altruistic, bal- 
anced, or crudely selfish rule before the representatives of absolutist 
power. Confucius pointed out that Yii, the legendary founder of the 
protohistorical Hsia dynasty, ate coarse foods, dressed poorly, dwelt 
in a modest house, and concentrated his energies on the irrigation 
canals. This great culture hero, whom Confucius considered flaw- 
less, 5 combined a minimum of personal demand with a maximum of 
public devotion. 

In the later period of China's early history the kings lived very 
comfortably; but the best among them are said to have sought a 
balance between their own and their subjects' interests. The philos- 
opher Mencius, who discussed this point, did not challenge the rul- 
ers' right to build lofty edifices, parks, and ponds by corvee labor; 
but he asked that the people be permitted to share these enterprises 
with their king. 6 

Thus the philosophers of ancient China assumed that within the 
framework of governmental needs there existed genuine alternatives 
for action. Without exception, however, the masters of the agrarian 
apparatus state satisfied the constructional, organizational, and ac- 
quisitive needs of their realm with a maximum stress on their own 
advantage and a minimum stress on the requirements of their sub- 

b. The Rulers' Managerial Optimum 

I n its early phase the hydraulic regime becomes stronger and wealth- 
ier with the growth of its hydraulic economy. But at a certain point 
the government can obtain additional revenue by intensifying its 

CHAPTER 4, C 129 

acquisitive rather than its productive operations. It is at this point 
that different power constellations lead to a different managerial 

The rulers' managerial optimum is maintained whenever the gov- 
ernment collects a maximum revenue with a minimum hydraulic 
effort. The people's managerial optimum is maintained whenever a 
maximum hydraulic achievement is accomplished with minimum adr 
ministrative expense. Intermediate arrangements involve the collec- 
tion of a large but not maximum revenue, a good part of which is 
used to produce sizable but not maximum hydraulic works. 

The rulers' responses to these alternatives show clearly the effect 
of total power on those who wield it. Beyond the zone of stimulating 
discrepancy, they generally push only those hydraulic enterprises 
that improve their own well-being; and they are most ingenious in 
developing new methods of fiscal exploitation. In short, they aim at 
the rulers', and not at the people's, managerial optimum. 

c. The Rulers' Consumptive Optimum 

Three major alternatives may also be distinguished in the sphere 
of consumption. The rulers' consumptive optimum is maintained 
whenever the masters of the hydraulic state arrogate to themselves 
a maximum of goods, which they may consume with a maximum of 
conspicuousness ("splendor"). The people's consumptive optimum is 
maintained whenever the nongovernmental members of society re- 
ceive a maximum of goods, which they may consume as conspicu- 
ously as they please. Intermediate arrangements to some degree favor 
the representatives of the government without, however, seriously re- 
stricting the quality or conspicuousness of popular consumption. 

Again the responses to these alternatives show the effect of total 
power on those who wield it. The proverbial splendor of Oriental 
despotism as well as the proverbial misery of its subjects have their 
roots in a policy that is directed toward the rulers', and not the peo- 
ple's, consumptive optimum. 

This optimum has both an economic and a legal aspect. By con- 
centrating the national surplus in their own hands, the rulers restrict 
the amount of goods physically available to nongovernmental con- 
sumers. By legally forbidding the general use of prestige-giving ob- 
jects, they reserve to themselves conspicuous consumption. In sim- 
pler hydraulic civilizations both aims can be achieved without much 
difficulty. Increasing social differentiations complicate matters, but 
they do not preclude a situation that, for all practical purposes, re- 
alizes the rulers' optimum. 


In the Inca empire the common people ate poorly and had little 
opportunity to drink heavily. 7 Their rulers ate extremely well, and 
they imbibed to excess. 8 Moreover, the gulf between the two groups 
was widened by laws which reserved the use of gold, silver, precious 
stones, colored feathers, and vicuna wool to the rulers. The common- 
ers were permitted some modest ornaments, but even these could be 
worn only on special occasions. 9 

Arrangements of this kind are most easily enforced when the great 
majority of the commoners are peasants living in government- 
controlled and more or less equalitarian villages. The emergence of 
many property-based enterprises involves the growth of nonbureau- 
cratic forms of wealth, both mobile and immobile; and such a de- 
velopment inevitably affects the pattern of consumption. 

Even under these circumstances the bulk of the rural and urban 
population continues to live poorly; and the small stratum of non- 
bureaucratic property-holders sees their fortunes constantly threat- 
ened by taxation and confiscation (and in time split up through the 
laws of inheritance). But wherever large property-based business be- 
came essential, private wealth could not be eradicated, and those 
possessing it could not be prevented from enjoying at least some 
part of it. 

Thus the laws which reserved certain types of dress or other con- 
spicuous goods to the ruling class became a crucial means for placing 
the men of the governmental machine and the priests of the domi- 
nant religion above the mass of the commoners. In traditional China 
the officials and their nonofficiating relatives were distinguished by 
their houses, furniture, clothes, and vehicles. 10 The Indian Law 
Books prescribe very precisely the garments, girdles, staffs, etc. to be 
used by Brahmins, Kshatriyas, and Vaisyas. 11 In the Near East dis- 
tinct bureaucratic features of dress are documented for Pharaonic 
Egypt, 12 Assyria, 13 Byzantium, 14 the Arab caliphate, 15 the Mamluks, 16 
and Ottoman Turkey. 17 

Within the limits of these regulations the commoners might — the- 
oretically speaking — enjoy their wealth. But they always hid their 
most precious possessions, and frequently their fear of confiscatory 
action was so great that they avoided all ostentation. The sweeping 
persecution of the merchants under the Earlier Han dynasty was 
provoked by the blatant show which the rich businessmen had made 
of their wealth. 18 Under a government which makes no effort to 
approach the rationality maximum, potential victims of confiscation 
may act with extreme caution. The French physician, Bernier, who 
from 1655 to ^58 lived in the Near East and afterward spent almost 
ten years in Mogul India, was struck by the frustrating atmosphere 

CHAPTER 4, C 131 

in which the businessmen of Asia operated. Enterprise found "little 
encouragement to engage in commercial pursuits," because greedy 
tyrants possessed "both power and inclination to deprive any man 
of the fruits of his industry." And "when wealth is acquired, as 
must sometimes be the case, the possessor, so far from living with in- 
creased comfort and assuming an air of independence, studies the 
means by which he may appear indigent: his dress, lodging, and 
furniture continue to be mean, and he is careful, above all things, 
never to indulge in the pleasures of the table." 19 

Bernier's observations must not be pressed. Under more far-sighted 
rulers the wealthy merchants of Asia lived luxuriously as long as 
their behavior did not invite disaster. And even in the India of 
Aurangzeb some few government-protected persons of wealth, Bernier 
tells us, "are at no pains to counterfeit poverty, but partake of the 
comforts and luxuries of life." 20 

But such exceptions do not negate the basic trend. In hydraulic 
civilizations wealthy commoners were denied the proprietary security 
which the burghers of the later Middle Ages enjoyed; and they did 
not dare to engage in the conspicuous consumption which the 
medieval businessmen practiced, despite the many sumptuary laws 
to which they too had to submit. The lavish display by the repre- 
sentatives of the state on the one side and the predominance of 
genuine and feigned poverty on the other spectacularly show the 
effect of total power on the consumptive optimum of hydraulic 

d. The Rulers' Judicial Optimum 

Similarly one-sided decisions characterize the judicial field. As 
explained above, no society is without standardized norms; and few 
advanced agrarian civilizations are without written or codified laws. 
Thus it is the special setting and intent that separate the laws of 
hydraulic despotism from those of pluralistically controlled states. 

The rulers' judicial optimum is maintained whenever the repre- 
sentatives of government exert a maximum influence on the formula- 
tion and application of their country's laws. The people's judicial 
optimum is maintained whenever the nongovernmental elements of 
society are decisive. In democratic commonwealths the constitu- 
tionally qualified citizen may participate in the formulation of the 
laws. He may exercise the functions of a judge, as he did in demo- 
cratic Athens, or he may, as a lay juror, cooperate with professionally 
trained, but elected judges. In both cases the nongovernmental forces 
of society, and not a despotic state, are charged with the application 


of the law. Intermediate variants are characterized by an increased, 
but not absolute, governmental power and by a proportionately de- 
creased popular control over the legislature and judiciary. 

It is obvious that the first type of judicial optimum prevails in 
hydraulic society. And it is equally obvious that in the judicial 
sphere, as in others, the masters of the hydraulic state seek a 
maximum of results (internal order) with a minimum of govern- 
mental effort and expense. This they accomplish not by yielding 
important judicial functions to quasi-independent secondary centers 
of power, as did the sovereigns of feudal Europe, 6 but by permitting 
politically irrelevant groups to handle certain of their own legal 
affairs, or by permitting magistrates to handle legal matters along 
with their other duties, or, where professional judges are the rule, 
by having as few full-time judges as possible. 

Such conditions preclude the development of independent juries. 
They discourage elaborate judicial procedures. And they leave little 
room for the functioning of independent professional lawyers. With 
these limitations the judges of a hydraulic society settle legal cases 
— many of which arise from clashes of proprietary interests, and in 
countries with a highly commercialized urban life this field of action 
may become very important indeed. 21 

However, even at their rational best, the laws of such countries 
express a fundamentally unbalanced societal situation. Even if they 
protect one commoner against the other, they do not protect the 
commoners — as individuals or as a group — against the absolutist 
state. Shortly after Bernier had commented on this phenomenon, 
John Locke did likewise; and his references to Ottoman Turkey, 
Ceylon, and Tsarist Russia show him aware that the tyrannical 
variant of judicial procedure, which English autocracy failed to de- 
velop fully, flourished unhampered under Oriental despotism. 

Locke insists that the presence of laws in a despotic regime proves 
nothing as to their justness: 

"if it be asked what security, what fence is there in such a state 
against the violence and oppression of this absolute ruler, the 
very question can scarce be borne. They are ready to tell you 
that it deserves death only to ask after safety. Betwixt subject 
and subject, they will grant, there must be measures, laws, and 
judges for their mutual peace and security. But as for the ruler, 
he ought to be absolute, and is above all such circumstances; 
because he has a power to do more hurt and wrong, it is right 

b. The holders of office land and the tax collectors who occasionally act as judges 
are, either fully or partially, integrated in the bureaucratic apparatus. See below, 
Chap. 8. 

CHAPTER 4, C 133 

when he does it. To ask how you may be guarded from harm 
or injury on that side, where the strongest hand is to do it, is 
presently the voice of faction and rebellion. As if when men, 
quitting the state of nature, entered into society, they agreed 
that all of them but one should be under the restraint of laws; 
but that he should still retain all the liberty of the state of 
Nature, increased with power, and made licentious by im- 
punity. This is to think that men are so foolish that they take 
care to avoid what mischiefs may be done them by polecats or 
foxes, but are content, nay, think it safety, to be devoured by 
lions. 22 

4. "Absolute Power Corrupts Absolutely" 

This is a bitter indictment. Contrary to modern apologists for 
totalitarian laws and constitutions, Locke refuses to put any trust in 
the autocrat's potential benevolence: "he that thinks absolute power 
purifies men's blood, and corrects the baseness of human nature, need 
read but the history of this, or any other age, to be convinced to 
the contrary." 23 Lord Acton's affirmative version of Locke's thesis 
is well known: "Power tends to corrupt and absolute power corrupts 
absolutely." 24 

Acceptance of this idea need not include an acceptance of Locke's 
pessimistic views on "the baseness of human nature." Man acts from 
many motives, which under different circumstances operate with 
different strengths. Both self-centeredness and community-centered- 
ness seek expression; and it depends on the cultural heritage and the 
over-all setting whether one or the other of them will prevail. A 
governmental — or proprietary — order leading to the emergence of 
absolute power encourages and enables the holders of this power to 
satisfy their own interests absolutely. It is for this reason that 
agrarian despotism, like industrial despotism, corrupts absolutely 
those who bask in the sun of total power. 

5. The Rulers' Publicity Optimum 

The corrupting influence is further consolidated by a one-sidedly 
manipulated public opinion. Public opinion may be shaped in a 
number of ways; and here, as elsewhere, the rulers' and the people's 
interests diverge sharply. This becomes clear as soon as the major 
alternatives are outlined. 

The rulers' publicity optimum is maintained whenever the gov- 
ernment's real or alleged achievements are given a maximum of un- 
critical publicity, while the people's experiences, sufferings, and 
views receive a minimum of notice. The people's publicity optimum 


combines a full presentation of the government's achievements and 
shortcomings. Intermediate arrangements favor the government 
without keeping the nongovernmental forces of society from stating 
their own case. 

Independent popular criticism differs both in quality and intent 
from the many and continued criticisms made by leading members 
of the officialdom. Bureaucratic criticism is vital to the proper 
functioning of complex administration, but it is voiced either behind 
closed doors or in publications accessible only to a limited number 
of educated persons, who are usually members of the ruling group. 
In both cases, the people's problems are viewed essentially from 
the standpoint of a more or less rationally conceived government 

Wielding total power, the masters of the hydraulic state can 
readily maintain the rulers' publicity optimum. Under socially un- 
differentiated conditions, the government's (frequently the sover- 
eign's) voice drowns out all criticism except as it may appear in such 
inconsequential media as popular tales and songs. More differ- 
entiated conditions provide additional outlets in secondary religions 
and philosophies, in popular short stories, novels, and plays. But 
even these media remain significantly feeble. In contrast to the in- 
dependent writers who, under Western absolutism, challenged not 
only the excesses but the foundations of the despotic order, the 
critics of hydraulic society have in almost every case complained only 
of the misdeeds of individual officials or of the evils of specific 
governmental acts.** Apart from mystics who teach total withdrawal 
from the world, these critics aim ultimately at regenerating a system 
of total power, whose fundamental desirability they do not doubt. 

6. The Two-fold Function of 
the Benevolence Myth 

a. It Stresses the Long-range Interest of the Despotic 

The advantages of the benevolence myth for the despotism which 
it glorifies are twofold. By presenting the ruler and his aides as 

c. In the total managerial societies of today, state-directed popular criticism is used 
to supplement and dramatize the government's criticism of problematic elements, 
particularly in the middle and lower echelons of the bureaucracy. Criticism of this 
kind has been encouraged in many hydraulic societies. The letters to Stalin differ 
technically, but not institutionally, from the letters and petitions addressed in the past 
to Oriental despots. 

d. Often government functionaries indict blundering fellow functionaries or harmful 
administrative procedures more sharply than do persons who are not part of the 

CHAPTER 4, C 135 

eager to achieve the people's rationality optimum, they enable the 
official spokesmen to educate and discipline the members of their 
own group. The holder of power, who operates below the rulers' 
rationality minimum, endangers the safety of the governmental 
apparatus, whereas one who operates above this level enhances the 
stability of the regime. He exploits his orchard as an intelligent 
gardener should. 25 Moreover, the ruler and his men must not weaken 
their position by crude managerial neglect, excessive taxation, or 
provocative injustice. The myth of an unselfish (benevolent) des- 
potism dramatizes these desiderata which, consciously or uncon- 
sciously, are underwritten by all thoughtful members of the ruling- 

b. It Weakens Potential Opposition 

More important still than the impact of the benevolence myth on 
the holders of power is its effect on the nongovernmental forces of 
society. The myth admits that individual sovereigns and officials 
may be unworthy, but it depicts the despotic order as fundamentally 
good— in fact, as the only reasonable and commendable system of 

Thus the embittered subject, who is permanently exposed to such 
propaganda, cannot well strive for the creation of a new and less 
despotic order. He and others who feel as he does may withdraw to 
the mountains. They may kill some local officials. They may defeat 
the government's men in arms. They may even overthrow a tottering 
dynasty. But eventually they will only revive — and rejuvenate — the 
agromanagerial despotism whose incompetent representatives they 
eliminated. The heroes of China's famous bandit novel, the Shui-hu 
Ch'uan, could think of nothing better to do than to set up on their 
rebel island a miniature version of the very bureaucratic hierarchy 
which they were so fiercely combating. 

c. The Presence of Good Sovereigns and Just Officials 
Fails to Upset the Prevailing Trend 

If man were exclusively self-centered, the result of all this would be 
very simple indeed. And very sad. But man is also community- 
centered. And this side of his character finds expression also in 
hydraulic society. To be sure, under the conditions of agrarian 
despotism, it is difficult to be a good sovereign or a just official. But 
it is not impossible. Throughout the hydraulic world serious- 
minded rulers attended to their managerial and judicial duties con- 
scientiously, and honest officials strove to prevent fiscal and judicial 
oppression. Courageous functionaries insisted on what they con- 


sidered proper policies, although by doing so they opposed the 
wishes of powerful superiors, and occasionally even of the sovereign 

But those who pursue such a course clash with the interest of the 
vast self-indulgent and scheming ruling group; and history shows 
that only a handful of unusually community-minded (ethically 
"possessed") persons was so disposed. Furthermore, even this pa- 
thetically small number of "good" men was not completely aware 
of how slanted the rulers' optimum was, which they recommended. 
Confucius' gentleman bureaucrat, the ideal ruler of the Bhagavad- 
gltdj and the "just" statesmen of the ancient Roman or Islamic Near 
East all try to be fair within the framework of a society which takes 
the patterns of despotic power, revenue, and prestige for granted. 

7. Hydraulic Despotism: Benevolent In Form, 
Oppressive in Content 

Thus agromanagerial despots may present their regimes as benev- 
olent; actually, however, and even under the most favorable cir- 
cumstances, they strive for their own, and not for the people's, 
rationality optimum. They plan their hydraulic enterprises accord- 
ing to what benefits their might and wealth. And they write their 
own ticket as fiscal masters of the national surplus and as conspicuous 

Stalin claims that in a modern industrial apparatus state the 
culture of a national minority is national in form and socialist in 
content. 26 Experience shows that the "socialist" (read: apparatchik) 
substance quickly wipes out all but the most insignificant national 
elements. A similar mechanism is at work in the agrarian apparatus 
state. Paraphrasing Stalin's formula and replacing myth by reality, 
we may truthfully say that hydraulic despotism is benevolent in form 
and oppressive in content. 



otal terror— total submission 
—total loneliness 


Man is no ant. His efforts to escape from freedom * show him am- 
bivalently attracted by what he ambivalently abandons. The urge to 
act independently is an essential attribute of homo sapiens, and a 
highly complex one. Not all of its components are socially valuable; 
but among them is man's most precious motivating force: the urge 
to obey his conscience, all external disadvantages notwithstanding. 

What happens to man's desire for autonomy under the conditions 
of total power? One variant of total power, hydraulic despotism, 
tolerates no relevant political forces besides itself. In this respect it 
succeeds on the institutional level because it blocks the develop- 
ment of such forces; and it succeeds on the psychological level, 
because it discourages man's desire for independent political action. 
In the last analysis, hydraulic government is government by in- 


i. The Need 

Man is no ant. But neither is he a stone. A policy that upholds the 
rulers' publicity optimum confuses the people's mind, without how- 
ever eliminating their feelings of frustration and unhappiness. Un- 
checked, these feelings may lead to rebellious action. To counter this 
dangerous trend the hydraulic regime resorts to intimidation. Terror 
is the inevitable consequence of the rulers' resolve to uphold their 
own and not the people's rationality optimum. 


138 total terror, submission, loneliness 

2. Its Official Recognition: 
"Punishment Is the King!" 

Many spokesmen of hydraulic despotism have emphasized the need 
for rule by punishment. Such a policy may be justified by the argu- 
ment that guiltless people are few. 1 Confucius preferred education 
to punishment; yet he, too, believed that it would take a hundred 
years of good government "to transform the violently bad and to 
dispense with capital punishment." 2 

Thus with varying arguments, punishment has been viewed as an 
essential tool of successful statecraft. The Hindu law book of Manu 
establishes fear-inspiring punishment as the foundation of internal 
peace and order. Punishment, which — of course — must be just, makes 
everyone behave properly. 3 Without it caste barriers would be 
crossed; and all men would turn against their fellows. "Where 
Punishment with a black hue and red eye stalks about," * subjects 
live at peace. "The whole world is kept in order by punishment." 5 

By punishment the ruler protects the weak against the strong, 
sacrifice against animal violation, property against its (nongovern- 
mental) enemies and social superiority against assaults from below. 
"If the king did not, without tiring, inflict punishment on those 
worthy to be punished, the stronger would roast the weaker, like 
fish on a spit: The crow would eat the sacrificial cake and the dog 
would lick the sacrificial viands, and ownership would not remain 
with any one, the lower ones would (usurp the place of) the higher 
ones." 8 Thus "punishment alone governs all created beings, punish- 
ment alone protects them, punishment watches over them while 
they sleep." 7 Indeed, "punishment is . . . the king." 8 

The rulers of ancient Mesopotamia claimed that they received 
their power from the great Enlil. 9 This terrifying god symbolizes "the 
power of force, of compulsion. Opposing wills are crushed and beaten 
into submission." 10 Although he is supposed to use his cruel might 
judiciously, 11 "man can never be fully at ease with Enlil but feels 
a lurking fear." 12 This being so, the sovereign's readiness to identify 
himself with Enlil or with deities descended from him is deeply 
significant. The Sumerian kings usually identified themselves with 
Enlil directly. 13 The Babylonians upheld the basic idea, but modi- 
fied it. Hammurabi pictured himself as having been "called" by 
Enlil; and he names Enlil's son, Sin, as his divine father. 14 In both 
cases the Mesopotamian rulers stressed the terroristic quality of their 

The terror inherent in Pharaonic despotism is symbolized by the 
poisonous Uraeus snake, which lies coiled on the ruler's forehead 
and threatens his enemies with destruction. 16 The king's actions are 

CHAPTER 5, B 139 

also compared with those of the fear-inspiring lion goddess, 

Chinese statecraft learned to express its need for terrifying punish- 
ment in the rational and moral form of Confucianism. But punish- 
ment was the primary weapon of the so-called Legalists and of such 
Legalist-influenced Confucianists as Hsiin Tsu. And it remained a 
cornerstone of official policy throughout the imperial period. What 
we would call the Ministry of Justice was known in traditional China 
as the Ministry of Punishments. 

The Islamic ruler saw to it that he was both respected and feared. 18 
The Arabian Nights, which depicts Harun al-Rashid usually ac- 
companied by his executioner, presents in fictional dress a historic 
truth. The executioner was a standard feature of the Abbassid 

3. The Morphology of Violence 

To be sure, all governments deserving the name have ways of im- 
posing their will on their subjects, and the use of violence is always 
among them. But different societies develop different patterns of 
integrating (or fragmenting) violence and of controlling (or not con- 
trolling) it. 

a. Integrated versus Fragmented Patterns of Violence 

In ancient Greece, free men ordinarily wore arms — according to 
Thucydides, "because their homes were undefended." 17 In other 
words, the government did not monopolize the use of force. With 
the growth of public safety the early custom disappeared in most city 
states; " but the citizens, who were potential warriors, were still per- 
mitted to keep the tools of violence in their homes. Pictorial evidence 
portraying the start of a campaign shows "mostly the woman bring- 
ing the weapons from the home to the departing man." 19 

In Medieval Europe the semi-independent feudal lords from the 
beginning represented important secondary centers of military action, 
and in the course of time many towns developed their own armed 
forces. These feudal and urban nuclei of political and military life 
were free to use violence both within their own jurisdictions and 
against one another. The vassal, who appeared before his sovereign 

a. See Breasted, 1927, I: 327, and cf. II: 92, and IV: 166; Erman, 1923: 78 ff.; and 
Wilson, 1950: 11. According to one story, Sekhmet emerged as the suppressor of a 
conspiracy. When the supreme god Re "perceived the things which were being 
plotted against him by mankind," he conjured up a force to crush the evil schemers. 
Then "Sekhmet came into being." She quickly "prevailed over mankind," and desiring 
to drink human blood — or what she believed to be human blood, "she drank, and it 
was good in her heart" (Wilson, 1950: 11). Cf. Erman, 1923: 78 ff. 


with his sword at his side, expressed strikingly the fragmented and 
balanced pattern of violence that characterized feudal society. 

Concentration of the legitimate uses of force in the hands of the 
state does not occur under conditions of total power only. Modern 
constitutional government restricts private violence more and more. 
But it differs from agrarian and industrial apparatus states in that 
the size, quality and use of coercion (army and police) are deter- 
mined by the nongovernmental forces of society. The experiences 
of classical Greece and the modern West show that a country may 
rally powerful armies without its citizens losing control over them. 

b. Controlled versus Uncontrolled Violence 

Army discipline requires unquestioning subordination; and the 
commander in chief of a well-coordinated army — which the feudal 
hosts were not — rules absolutely within the limits of his jurisdiction. 
However, in a democratic country he remains responsible to the 
citizens who control the government. General Eisenhower's com- 
ments on the Soviet method of attacking through mine fields indicate 
the institutional alternatives. In "a matter-of-fact statement" Marshal 
Zhukov explained to the American general: "When we come to a 
mine field our infantry attacks exactly as if it were not there. The 
losses we get from personnel mines we consider only equal to those 
we would have gotten from machine guns and artillery if the Ger- 
mans had chosen to defend that particular area with strong bodies of 
troops instead of with mine fields." Eisenhower adds drily: "I had 
a vivid picture of what would happen to any American or British 
commander if he pursued such tactics, and I had an even more 
vivid picture of what the men in any one of our divisions would 
have to say about the matter had we attempted to make such a 
practice a part of our tactical doctrine." 20 

The Soviet way saves materiel and time; and it suits to perfection 
the rulers' tactical optimum. Obviously this optimum can be realized 
only when organized violence is wielded by the masters of an un- 
checked state. The social quality of organized violence, like that of 
other governmental functions, changes with the over-all setting in 
which it develops. 


The subjects of an agrarian apparatus state have little opportunity 
to argue the problem of uncontrolled violence. They may be per- 
mitted the possession of small and simple weapons, particularly in 
the villages, which have to ward off bandits. But the organized and 

CHAPTER 5, <2 141 

military use of coercion is essentially concentrated in the hands of 
the absolutist rulers, who usually give audience only to unarmed 
men. In hydraulic society the monster with "a black hue and red 
eye" is no watch-dog tied up by the people, but a tiger that moves 
at will. 

1. Its Physical Aspect 

Like the tiger, the engineer of power must have the physical means 
with which to crush his victims. And the agromanagerial despot does 
indeed possess such means. He exercises unchecked control over 
the army, the police, the intelligence service; and he has at his 
disposal jailers, torturers, executioners, and all the tools that are 
necessary to catch, incapacitate, and destroy a suspect. 

2. Its Psychological Aspect 

a. Unpredictability 

Furthermore, he can employ these devices with maximum 
psychological effect. Everywhere persons wielding great govern- 
mental or proprietary power like to shroud certain of their acts in 
secrecy; but the procedures of a despotic government are enigmatic 
because of the very nature of the regime. Accountable only to them- 
selves, the men of the apparatus tend to handle even insignificant 
matters with secretiveness; and they raise mystification to an art 
when they want to intimidate and surprise. Unpredictability is an 
essential weapon of absolute terror. 

b. Lenin: ". . . power not limited by any laws" 

Lenin defined the dictatorship of the proletariat — which he held 
to be the heart of the Soviet regime — as "a power not limited by any 
laws." x Like other utterances of Lenin, this formula combines an 
impressive half-truth with important fallacies. First, the Soviet 
dictatorship was never controlled by the Russian workers; and there 
is ample evidence that Lenin knew this. Second, no regime, however 
dictatorial, operates without normative regulations or laws of some 
kind; and this, too, was well known to Lenin. Before he made the 
just-quoted statement, his dictatorial government had already issued 
many revolutionary statutes and decrees. 2 The despot's right to in- 
terpret, change, and override previously established laws is a funda- 
mental constitutional and legal principle of absolutist rule. Lenin's 
definition stresses with brutal frankness the dictator's unchecked 
power to use laws as he wishes. In the sphere of terror he may go 


so far that it becomes difficult to distinguish between lawless terror 
and terror by law. 

c. Lawless Terror and Terror by Law 

A chief or ruler does not necessarily override the laws of his 
hydraulic community when he himself commits — or gives orders to 
commit — acts of terrifying brutality. 

In smaller hydraulic tribes autocratic cruelty is no issue, because 
the chief, being close to his fellow tribesmen, is unable to exert 
power over and above his directing functions. This is the case among 
the Suk and their hydraulic neighbors and throughout the Ameri- 
can Pueblos. 

In larger hydraulic tribes the chief may seek to bolster his incipient 
autocracy by the employment of spectacular terror. A Chagga chief, 
for instance, may commit all manner of cruelties against his sub- 
jects. Ndeserno is said to have torn the hearts from his victims' 
bodies while they were still alive and to have had them roasted for 
his children. 3 A chieftain who went to such extremes was con- 
templated with grave apprehension, but, according to Gutmann, 
"such cruelties against individuals did not harm his prestige." On 
the contrary, the fear they inspired cemented the stability of the 

The spectacular terror directed by the rulers of ancient Hawaii 
may well have served the same purpose; 5 and the so-called Cannibal 
Texts of the Old Kingdom suggest a similar situation in prehistoric 
Egypt. One of these texts, found in a pyramid, reveals a dead ruler 
killing, dissecting, and cooking human beings in the nether world for 
his gustatory pleasure; 6 and another reveals him as taking "the 
wives from their husbands whenever he wants to and according to his 
heart's desire." & 

In more differentiated hydraulic civilizations, there is less need to 
bulwark the ruler's exalted position by spectacular acts of autocratic 
ruthlessness. Although such acts do not completely cease, they are 
now initiated mainly by excessively cruel (and/or insecure) sover- 
eigns and by the heads of dynasties which operate below the rulers' 
rationality maximum. Gaudefroy-Demombynes describes the irra- 
tionally terroristic quality of the Abbassid caliphate as follows: 
"Improvised executions and the exhibition of heads are part of the 
regular life of the Abbassid court. Beginning with the reign of El 
Manc,our, when a person is urgently summoned to the palace by 
the guards of the caliph, he feels that he has a good chance not to 

a. Sethe, PT, II: 354 ft.. The Chagga chiefs seem to have made a like claim on all 
girls and women of their realm (Widenmann, 1899: 48; cf. Gutmann, 1909: 25). 

CHAPTER 5, C 143 

return alive. He makes his testament, says farewell to his family, and 
carries his shroud under his arm." b 

In these and other instances, the ruler's terroristic behavior was 
above rather than against the law. On the other hand, officials who 
resorted to extreme brutalities often went beyond even the broadest 
possible interpretation of the law. At times they might be held 
accountable. But many "lawless" bureaucratic terrorists were criti- 
cized only after they were dead. 

The excesses of autocratic and bureaucratic terror are an extreme 
manifestation of human behavior under total power. Institutionally, 
however, they are probably less important than the innumerable acts 
of terror that were perpetrated as a matter of routine and within 
the flexible frame of despotic law. It was this routine terror in 
managerial, fiscal, and judicial procedures that caused certain ob- 
servers to designate the government of hydraulic despotism as 
"government by flogging." 

3. "Government by Flogging" 

a. Terror in Managerial Procedures 

"The language of the whip" seems to have been employed 
regularly in the state corvees of ancient Sumer. 7 Under the Pharaohs, 
every government administrator could resort to corporal punish- 
ment. 8 The pictorial records of ancient Egypt show men conducting 
all manner of public enterprises with sticks in their hands. 9 In the 
later part of the 19th century, when the British began to abolish 
"government by flogging," the whip was still standard equipment for 
insuring the success of the hydraulic corvee. 10 Present-day writers who 
are greatly impressed by the planned economy of the Incas would 
do well to remember that the Inca prince, Garcilaso de la Vega, 
glorying in his forebears' achievements, took it for granted that the 
one sure way to make people industrious was to threaten them with 
beating. 11 

b. Terror in Fiscal Procedures 

Since the days of the Pharaohs, reluctance in paying taxes was 
overcome by force. A famous satire of the New Kingdom tells that 
the Egyptian peasant who failed to deliver his quota of grain was 
"beaten, tied up, and thrown into the ditch." 12 Irregularities in 

b. Gaudefroy-Demombynes, 1931: 384. The friend of an Abbassid caliph, who went 
to the court every Friday, was "gripped by an intense fear" when he was summoned 
on a different day. Had he been maligned? Had he been found wanting? His "anguish 
and fear" increased until he discovered to his immense relief that the sovereign 
merely wanted him to share an hour of idleness and pleasure (Sauvaget, 1946: 62). 


handling state and temple property also called for corporal punish- 
ment. 1 * 

The Sacred Law of Islam prohibited torture; but the tax officials 
of the caliphs apparently found it impossible to fulfill their task 
without resorting to violence. 14 Under the Abbassid dynasty, torture 
was a concomitant of tax gathering until the year 800; and after a 
short interlude of about twelve years it was invoked again, and as 
brutally as ever. Government agents "beat the people, imprisoned 
them, and suspended heavy men by one arm so that they almost 
died." » 

The Arthashastra made it mandatory for police and court judges 
to see that rural taxes were duly paid, and to use force if necessary. 16 
The Law Code of imperial China prescribed beating as the standard 
punishment for persons who failed to fulfill their fiscal obligations. 17 

c. Terror in Judicial Procedures 

The Chinese Code carried the issue of violence beyond the spheres 
of fiscal action. In case of continued resistance and/or inability to 
deliver, the defaulter might be taken before a judge; and if necessary, 
fiscal terror might be replaced by judicial terror. Judicial torture to 
extort evidence — and frequently also to punish — was employed in 
virtually all hydraulic civilizations. 

In Pharaonic Egypt beating was a regular adjunct of judicial 
procedures. 18 "He was examined with the rod" was standard phrasing 
in the New Kingdom. 19 

Indian, Chinese, and Islamic sources describe judicial terror in 
considerable detail. The Arthashastra states that "Those whose guilt 
is believed to be true shall be subjected to torture." 20 With the 
exception of the Brahmins," they could be given the "six punish- 
ments," the "seven kinds of whipping," the "two kinds of suspension 
from above," and the "water-tube." 21 Regarding persons "who have 
committed grave offences," the famous book is still more specific. 
They could be given the 

nine kinds of blows with a cane: 12 beats on each of the thighs; 
28 beats with a stick of the tree (nakta-mala); 32 beats on, each 
palm of the hands and on each sole of the feet; two on the 
knuckles, the hands being joined so as to appear like a scorpion; 
two kinds of suspensions, face downwards (ullambane chale); 
burning one of the joints of a finger after the accused has been 

c. They could not be tortured to extort evidence; but if found gu»Hy of 3 very 
grave crime, they could be branded (Arthflfastra, 1923: 270). 

CHAPTER 5, C 145 

made to drink rice gruel; heating his body for a day after he has 
been made to drink oil; causing him to lie on coarse grass for 
a night in winter. These are the 18 kinds of torture. . . . Each 
day a fresh kind of the torture may be employed. 22 

In particularly serious cases, such as attempts to seize the king's 
treasury, the accused could be "subjected once or many times to one 
or all of the above kinds of torture." 2S 

The Chinese Law Code describes a number of instruments used 
to extract evidence; 2i and the writings of sincere administrators 
elaborate on proper and improper methods of torture. 28 

Canonic prohibitions notwithstanding, the secular courts of the 
caliphs extorted evidence by employing "the whip, the end of a rope, 
the stick, and the strap on the back and belly, on the back of the 
head, the lower parts of the body, feet, joints, and muscles." 26 

Similar methods seem to have persisted in the Near East until 
recent days. In 19th-century Egypt, "justice, such as it was, was 
almost as much a terror to the innocent witness as to the accused 
person against whom testimony was borne." 27 

d. Western Correspondences Noteworthy for Their 
Temporary Strength and Their Limitations 

Manifestly, judicial torture is widespread in the hydraulic world. 
But is it specific? After all, torture had a definite place in Roman 
law. It appears prominently in late feudal and postfeudal Western 
legal procedures and in the Inquisition. And it survives today in the 
third degree. 

All these phenomena must indeed be recognized for what they 
are. They remind us grimly that human nature is the same every- 
where and that man succumbs to the corrupting influence of power 
whenever circumstances permit. Fortunately, the shape of Western 
institutions kept these inclinations from asserting themselves last- 
ingly. But the momentum they gained at certain times and in 
certain places precludes the complacent assumption that what hap- 
pened under hydraulic governments — and what is happening today 
in the totalitarian states — cannot happen here. 

The indigenous free men of ancient Greece and republican Rome 
did not employ managerial or fiscal terror against their fellow citi- 
zens — the citizens did not render corvee service nor did they pay 
substantial taxes — and "as a rule" they were not subjected to judicial 
torture. 28 Their societal order was too balanced for this; yet it was 
not sufficiently balanced to prevent the use of managerial and judicial 
terror against certain alien and unfree elements. In Greece, the 


position of most slaves was "not much different from that of domestic 
animals." 29 Their masters were free to punish them physically; 30 
and the not too numerous state slaves occupied in public works 
were directed by foremen, who, frequently slaves themselves, "had 
a name for being very hard." 31 In Greece both slaves and free aliens 
were the targets of judicial torture. 32 In republican Rome only slaves 
were so treated. 33 

The crystallization of absolutist power under the empire deprived 
the Roman citizens of the protection which their forefathers had 
enjoyed against judicial and other forms of governmental terror. 
Roman law in late Roman and Byzantine times extended judicial 
torture to the bulk of ail free persons. 34 

A similar change occurred in the later part of the Middle Ages. 
Early Frankish (Salic) law permitted only persons of servile status 
to be tortured. 35 Conflicts between free men were handled by courts 
composed of peers. Serious legal issues were settled by ordeal or 
judicial combat; 36 and the burghers in medieval towns, who orig- 
inally followed these procedures, soon preferred more humane and 
rational methods of determining guilt or innocence. 37 

The introduction of judicial torture — significantly bulwarked by 
references to Roman law — coincides with the rise of centralized and 
despotic power on a territorial and national scale. 38 Most historians 
point out that the procedures of the absolutist courts superseded 
the feudal methods of ordeal and combat.** Less frequently do they 
mention the equally important fact that the new judicial torture also 
replaced the significant beginnings of rational judicial procedure 
developed in the burgher-controlled towns. e 

Changes in judicial procedures were certainly intensified by the 
Inquisition; and anyone who studies this period is struck by the 
elaborate and cruel tortures employed in questioning heretics. How- 
ever, three points deserve attention: First, the Church, which based 
itself on medieval Canonic Law, did not originally recommend the 
use of extreme measures against heretics. 39 Second, judicial torture 
was probably initiated by secular agencies/ Third, terroristic proce- 
dures were equally harsh under those absolutist governments of 
Europe which, in the course of the Reformation, had dissociated 

d. Cf. Petit-Dutaillis, 1949: 309; Lea, i8g2: 480, 487 ff., 500 ff., 505. Lea describes in 
some detail what he calls the "resistance of feudalism" to the development of judicial 
torture (1892: 494 ff.). See also Williams, 1911: 72. 

e. In the 14th century the Italian communities continued to combat the increasing 
use of torture (Lea, 1892: 506 ff.); and in Liibeck, Germany's foremost city of burgher 
independence, legal orders discouraging ordeal, judicial duel, and torture yielded but 
slowly to the new absolutist law (ibid.: 483). 

f. Lea, 1908, I: 221; cf. Guiraud, 1929: 86. In the 12th century, long before judicial 
torture was institutionalized, heretics had been tortured to death (Helbing, 1926: 106 ff.). 

CHAPTER 5, C 147 

themselves from Rome. 40 No doubt the disintegration of medieval 
society stimulated both heretic tendencies and the fanatic desire to 
eradicate them; but it was only within the framework of rising 
absolutist state power that this desire took the form of the Inquisi- 

The limitations of Western absolutism also determined the point 
beyond which the representatives of despotic power could not subdue 
their own subjects. For a time they were able to employ judicial 
terror in secular and religious matters, but managerial and fiscal 
terror were not invoked against the bulk of the population. With the 
rise of modern industrial society judicial torture was eliminated in 
the heartlands of European absolutism, and eventually also in the 
terror-ridden slave economy of our southern states. Presently, public 
opinion is crusading against such police actions as the third degree. 
These methods were never legal; their illegal use is receding before 
the growing vigilance and strength of public-minded citizen organiza- 

Pre-Mongol ("Kievan") Russia accepted many elements of By- 
zantine law, but not the use of corporal punishment. This device, 
as well as judicial torture, seems to have emerged in Russia only 
when an Oriental type of despotism arose during and after the Tatar 
period.* 1 Third degree methods continued to be employed until the 
last decades of the Tsarist regime; 42 but torture as a means of 
getting evidence was discarded early in the 19th century, when the 
growth of property-based industrial forms of life promoted the 
restriction of many absolutist features of Russian law and society.' 
It was left to the masters of the Communist apparatus state to reverse 
the humanizing trend and to reintroduce the systematic infliction of 
physical pain for the purpose of extracting "confessions." h 

4. Varying Configurations of Terror in the 
Hydraulic World 

a. Relatively Lenient Developments 

In different areas and phases of the hydraulic world the methods of 
terror differed. The indigenous Babylonian government, for in- 

g. Lea, 1892: 581; Williams, 1911: 79. For occasional late occurrences see Williams, 
loc. cit., and Scott, 1943: 264. George Kennan, who at the close of the 19th century, 
studied the life of political prisoners and exiles in Siberia, draws attention to the 
arbitrary methods employed by the Tsarist police: unjust arrests and imprisonment, 
beating and torturing (Kennan, 1891, II: 52 ff.). These methods were certainly brutal, 
but the growing strength of public opinion restricted them increasingly; and a com- 
parison of the conditions described by Kennan and those to which Soviet prisoners 
are subjected today reveals an abysmal retrogression in judicial procedure. 

h. The Communist methods of judicial terror vary with time, space, circumstance, 


stance, proceeded close to the rulers' rationality maximum; and 
Babylonian laws known to us mention, as means of establishing guilt 
or innocence, the ordeal, the oath, and witnesses, but not torture. 43 
To be sure, judicial torture may well have been employed in cases 
involving the security of the regime (the Code does not discuss these 
matters); even for minor offenses against the interests of the govern- 
ment punishment was terrify ingly harsh; * and there is no reason 
to assume that the "language of the whip," which accompanied the 
Sumerian corvee, was not used by Babylonian master builders and 
master irrigators. But while the Babylonian state, local administra- 
tive councils notwithstanding, remained an absolutist regime, it acted 
as rationally in judicial and many other matters as could be expected 
under the conditions of an agromanagerial system of total power. 

b. Average and Excessive Developments 

In most hydraulic civilizations the rulers employed fully all major 
forms of terror, the managerial, the fiscal, and the judicial. In doing 
so, they established procedural averages, which occasionally were 
codified. These averages usually sufficed to satisfy the needs of the 
regime; but not infrequently those who applied them resorted to 
methods of extreme brutality, which besides producing quicker re- 
sults, yielded a surplus income for the officials who perpetrated 

As shown above, not all officials went to such lengths; and for 
various reasons extreme malpractice might be punished. But 
"moderate" excesses tended to remain unchallenged. And from 

and purpose; but despite a certain ingenuity in applying psychological devices, the 
main techniques can hardly be claimed as inventions. The "keeping-awake" torture, a 
seemingly mild but actually irresistible way of breaking the will of a person under 
interrogation, appeared in the Roman arsenal of planned cruelty under the name 
tormentum vigiliae (Helbing, 1926: 45). It was re-"invented" in 153a by Hippolytus de 
Marsiliis (Williams, 1911: 77). The starvation torture was known as tormentum famis 
(Helbing, 1926: 4k). Certain Communist methods parallel procedure used by the 
Inquisition. Compare the abrupt changes from bad to good treatment and from 
good treatment to bad, and the facing of the prisoner with confessions or alleged 
confessions of others (Lea, 1908, I: 415 ff.). Cruder methods of torture, beginning with 
simple beating — Roman forerunner: the verbera (Helbing, 1926: 45) — attain their goal 
faster than the more "cultivated" tormentum vigiliae. They seem to be extensively 
employed particularly in times of crisis, such as the Great Purge, World War II, and 
the period of continued stress that followed this war (see Beck and Godin, 1951: 53 ff.; 
Weissberg, 1951: 238 ff., 242, 246, 296; SLRUN, 1949: 56, 67, 74 ff.). Of course, many 
Soviet modes of torture were foreshadowed by Ivan IV and his successors. 

i. Stealing government or temple property was punished with death (Hammurabi, 
Sees. 6, 8. See also translator Meek's note js. 

CHAPTER 5, D 149 

the standpoint of the commoner, the despotic apparatus remained 
irrationally formidable even when it employed only the standard 
methods of terror. It became frightening when it exhausted its 
terroristic potential. 


1. Man's Response to the Threat of Total Terror 

a. The Postulate of Common Sense and the Virtue of 
Good Citizenship: Obedience 

Living under the threat of total terror, the members of a hydraulic 
community must shape their behavior accordingly. If they want to 
survive, they must not provoke the uncontrollable monster. To the 
demands of total authority common sense recommends one answer: 
obedience. And ideology stereotypes what common sense recom- 
mends. Under a despotic regime, obedience becomes the basis of 
good citizenship. 

Of course, life in any community requires some degree of coor- 
dination and subordination; and the need for obedience is never 
completely lacking. But in the great agrarian societies of the West 
obedience is far from being a primary virtue. 

In the democratic city states of ancient Greece the good citizen 
was expected to display four major qualities: military courage, 
religious devotion, civic responsibility, and balanced judgment. 1 
Prior to the democratic period, physical strength and courage were 
particularly valued. 2 But neither the Homeric age nor the classical 
period considered unquestioning obedience a virtue in a free man, 
except when he served in the army. Total submission was the duty — 
and the bitter fate — of the slave. The good citizen acted in accordance 
with the laws of his community; but no absolute political authority 
controlled him absolutely. 

Nor did the loyalty which the medieval knight owed his overlord 
result in total submission. The feudal contract bound him to follow 
his sovereign only in a qualified and limited way. Among the virtues 
of the good knight, good horsemanship, prowess in arms, and courage 
ranked high. 3 Unquestioning obedience was conspicuously lacking. 

In hydraulic society the relation between the ordinary members 
of the community and their leaders was regulated very differently. 
The quest for integrated subordination appears even at the tribal 
level. In the American Pueblos submissiveness and a yielding dis- 
position are systematically cultivated. 4 Among the Chagga, "respect 


for the chief is the first command, which the parents impress upon 
their children." 5 

In state-centered hydraulic civilizations the supreme holders of 
power are not as close to the people as they are in Pueblo society, nor 
are they, as in certain Pueblos and among the Chagga, restrained 
by clan influence. The masters of an agrarian apparatus state make 
greater demands than the Pueblo leaders; and their means for en- 
forcing their will far surpass the modest political devices of Chagga 

Thorkild Jacobsen, discussing society and religion in ancient 
Mesopotamia, lists obedience as the prime virtue. Essentially "in 
Mesopotamia the 'good life' was the 'obedient life.' "° Unlike the 
warriors of Medieval Europe, who often fought in small bands and 
with little concern for a ranking leader, the Mesopotamians felt that 
"soldiers without a king are sheep without their shepherd," "peas- 
ants without a bailiff are a field without a plowman," and "work- 
men without a foreman are waters without a canal inspector." 7 Thus 
the subject was expected to carry out the orders of his foreman, his 
bailiff, and — of course — his king. "All these can and must claim 
absolute obedience." 8 Submission which cannot be avoided is con- 
veniently rationalized: "The Mesopotamian feels convinced that 
authorities are always right." 9 

Similar concepts can be found in Pharaonic Egypt. A ship must 
have its commander, a gang its leader; 10 and whoever wants to sur- 
vive — and to succeed — must fit himself into the edifice of superordi- 
nation and subordination: "Bow thy back to thy superior, thy over- 
seer from the palace [the government]. . . . Opposition to a supe- 
rior is a painful thing (for) one lives as long as he is mild." " 

The law of Hindu India prescribes subordination to both secular 
and priestly authority. Those who oppose the king's commands suffer 
"various kinds of capital punishment." l2 

The Koran exhorts believers to obey not only Allah and his 
prophet but also "those in authority amongst you." 1S In the ab- 
solutist states established by Mohammed's followers, this passage was 
invoked to emphasize the basic importance of obedience in main- 
taining governmental authority. 14 

Confucius envisioned an authority that would realize the ruler's 
rationality maximum. He therefore insisted that every official should 
judge the propriety of the ruler's actions; and when conflict became 
serious, a top-ranking minister might retire. 15 Normally, however, 
the ideal functionary obeyed his ruler; 1C and reverence toward a 
superior was a basic duty. 17 The commoner was given no choice 
whatsoever. Since he could not understand the issues involved, he 


had to be "made to follow" what superior authority and insight 
dictated. 18 In Confucius' good society, as in its Indian and Near 
Eastern variants, the good subject was the obedient subject. 

2. Preparation for Total Obedience: 
Disciplinary Education 

The good subject was also the obedient son. For Confucius an 
education that demands absolute obedience to parent and teacher 
forms the ideal foundation on which to build absolute obedience to 
the masters of society. 

No similar correlation can be established for Medieval Europe. 
The son of a feudal knight was mercilessly disciplined. At an early 
age he was compelled to ride a high horse, while tied to the saddle; 
and to toughen him further he was buried in horse manure. 19 Curses 
and blows were frequent accompaniments to growth. Feature for 
feature, the early education of the young feudal knight seems to 
have been as harsh, or harsher, than the education of the young son 
of an Oriental official. And the apprenticeship of the young Euro- 
pean craftsman was no bed of roses either. 20 

But the behavior of the young burghers on festive occasions 
showed that the educational disciplines to which they had been ex- 
posed were not seriously inhibiting, 21 and the behavior of young 
knights remained equally carefree. Both groups matured under con- 
ditions that were built on contractual relations rather than on ab- 
solute authority, and they took their early frustrations as the passing 
experience that it actually was. 

Conversely, similar — or even less harsh — disciplines may be emi- 
nently effective for assuring total submission. In ancient Mesopo- 
tamia, "the individual stood at the center of ever wider circles of 
authority which delimited his freedom of action. The nearest and 
smallest of these circles was constituted by authorities in his own 
family: father and mother, older brother and older sister." 22 And 
"obedience to the older members of one's family is merely a begin- 
ning. Beyond the family lie other circles, other authorities: the 
state and society." Each and every one of them "can and must claim 
absolute obedience." 23 

The wisdom of ancient Egypt consciously interlinks obedience 
at home to obedience to the official. The obedient son "will stand 
well in the heart of the official, his speech is guided with respect to 
what has been said to him." 2i In Hindu India the demand for sub- 
ordination to the secular and priestly authorities is reenforced by 
the demand for subordination in the personal spheres of life. Obedi- 


ence is particularly due "the teacher, the father, the mother, and an 
elder brother." 2B 

Confucianism describes filial piety as a unique preparation for 
civic obedience: "There are few who, while acting properly toward 
their parents and older brothers, are inclined to oppose their su- 
periors. And there is nobody who, while averse to opposing his 
superiors, is inclined to making a rebellion." 26 

3. The Great Symbol of Total Submission: 

Education teaches man to obey without question, when despotic 
authority so demands. It also teaches him to perform gestures of 
reverence when the symbol rather than the submissive action is re- 
quired. True, all cultures have ways of demonstrating respect; and 
many gestures indicate subordination. 27 But no symbol has expressed 
total submission as strikingly, and none has so consistently accom- 
panied the spread of agrarian despotism, as has prostration. 

Total submission is ceremonially demonstrated whenever a sub- 
ject of a hydraulic state approaches his ruler or some other repre- 
sentative of authority. The inferior man, aware that his master's 
wrath may destroy him, seeks to secure his good will by humbling 
himself; and the holder of power is more than ready to enforce and 
standardize the symbols of humiliation. 

The inferior person may indicate his submissiveness by placing 
one hand over the other, as if they were tied together. 28 He may raise 
his open hands as a gesture of self-disarmament. Or going to ex- 
tremes, he may fall forward on all fours like an animal, strike his 
head on the ground, and kiss the dust. Under the shadow of Oriental 
despotism, prostration is an outstanding form of saluting the sov- 
ereign or other persons of recognized authority. The details vary; 
and occasionally symbols with similar intent are used. Generally 
speaking, however, prostration is as characteristic for hydraulic so- 
ciety as it is uncharacteristic for the higher agrarian civilizations 
of classical antiquity and the European Middle Ages. 

The absence of prostration in primitive hydraulic societies indi- 
cates the limitations of chiefly authority under tribal conditions. 
The Pueblo Indians hold their cacique in the highest esteem; but 
there are no evidences of the demonstrative submission that found 
open expression in the higher hydraulic civilizations of Aztec 
Mexico or Inca Peru. The Chagga tribesmen hail their chieftain; 

a. 0strup, 1929: 28 ff. Cf. the modern "hands up." 

CHAPTER 5, D 153 

and they murmur respectfully when he arrives or rises. 29 But this 
apparently is as far as their display of deference goes. 30 

In state-centered hydraulic civilizations prostration occurred al- 
most everywhere. In ancient Hawaii political power was sufficiently 
terrifying to make the commoners crawl before their rulers. 6 In 
Inca Peru, even the highest dignitary approached his sovereign like 
a bearer of tribute, his back bent under a load. 31 In pre-Conquest 
Mexico supreme reverence was expressed by prostration. Taught in 
the "colleges," 82 it was performed before royalty, men of distinc- 
tion, 33 and persons believed to be divine. 34 

In China prostration was practiced from the early days of the 
Chou dynasty — that is, during the pre-empire period of the terri- 
torial states; 35 and it prevailed throughout all subsequent phases of 
Chinese history. The experiences of the European envoys, who were 
asked to kowtow before the Manchu emperor, reveal both the im- 
portance of the custom and the embarrassment it caused Western 

In the classical days of Hindu India great respect was shown by 
embracing a person's feet; and the king seems to have been ap- 
proached in an attitude of prayer. 36 Prostration was performed be- 
fore deities and the teacher's young wife. However, in the later part 
of the Hindu period, the prime gesture of total submission was 
also performed before the sovereign. 37 Under Muslim rule both the 
sovereign 3S and venerable Hindus 39 were so honored. 

The importance of prostration in the Near East can be amply 
documented. The records of Pharaonic Egypt describe the whole 
country as "prone upon the belly" before a representative of the 
king. 40 Faithful subordinates are shown crawling, and kissing (or 
sniffing) the monarch's scent. 41 Pictorial evidence suggests that in 
the New Kingdom high dignitaries employed other gestures of 
reverence; 42 but contemporary sources do not say that they ceased 
prostrating altogether. They indicate clearly that lowly persons and 
subject peoples continued to prostrate. 43 

In ancient Mesopotamia prostration was performed before the 
gods, the ruler, and other distinguished personalities, 44 and it was 
performed also in Achaemenian Persia. 45 It persisted in the Hel- 

b. Fornander, HAF, VI: 12, 34 (religious prostration), 26 (before the king's idol); 
prostration before ruler: Kepelino, 1932: 12; Alexander, 1899: 26 ff.; Blackman, 1899: 


c. Cf. Manu, 1886: 69. In the second case, prostration obviously was performed in 

order to prevent bodily contact. For religious prostration, see Jatakam, III: 284; IV: 231; 
V: 274; VI: 302. 


lenistic empires of the Seleucids 4e and the Ptolemies, 47 and also in 
Sassanid Persia. 48 It became the standard gesture of reverence in 
Eastern Rome on the eve of the Byzantine period. 49 Needless to say, 
it fitted the social climate of Byzantium to perfection. 50 

The followers of Mohammed originally prostrated only in prayer. 
Eventually, however, the "Orientalized" Arabs, like the Greeks be- 
fore them, prostrated also in secular life. 51 In Ottoman Turkey the 
practice prevailed until close to the end of the Sultanate.* 

Thus in the hydraulic world prostration was the outstanding ex- 
pression of submission and reverence. Occasionally, equivalent ges- 
tures were used for the same purpose; and in a number of cases 
prostration spread to countries that were not ruled by Orientally 
despotic governments. However, the fate of the proskynesis in Medi- 
eval Europe shows how difficult it was to force this humiliating salu- 
tation on a politically balanced society. Some rudiments of the By- 
zantine ceremony survived in the ceremonial of the Western 
Church; yet the attempt of certain Carolingian rulers to uphold it 
as a secular ritual did not succeed. In Sicily under Roger II and 
Frederick II prostration was practiced temporarily probably under 
the influence of the Byzantines, 52 or the Arabs, who immediately 
preceded the Norman rulers. 53 

No doubt usage dulled man's sensitivity to the humiliating intent 
of prostration, and aesthetic accomplishment sweetened perform- 
ance. But no matter how much prostration was rationalized, it re- 
mained through the ages a symbol of abject submission. Together 
with managerial, fiscal, and judicial terror, it spectacularly marked 
the range — and the total power — of agrarian despotism. 


1. Loneliness Created by Fear 

Demonstrative and total submission is the only prudent re- 
sponse to total power. Manifestly, such behavior does not gain a 
superior's respect; but other ways of proceeding invite disaster. 
Where power is polarized, as it is in hydraulic society, human rela- 
tions are equally polarized. Those who have no control over their 
government quite reasonably fear that they will be crushed in any 
conflict with its masters. 

And the formidable might of the state apparatus can destroy not 
merely objectionable nongovernmental forces — with equal thor- 
oughness it may also overwhelm individual members of the ruling 

d. 0strup, 1929: 32; Lane, 1898: 211 (kissing the feet as a sign of abject submission). 

CHAPTER 5, E 155 

group, the ruler himself included. Many anxieties darken the path 
of life; but perhaps none is as devastating as the insecurity created by 
polarized total power. 

a. The Ruler: Trust No One! 

The ruler, being most illustrious, is also most to be envied. Among 
those near him, there are always some who long to replace him. 
And since constitutional and peaceful change is out of the question, 
replacement usually means one thing and one thing only: physical 
annihilation. The wise ruler therefore trusts no one. 

For obvious reasons the innermost thoughts of despots have been 
little publicized. But observable behavior and utterances confirm 
our assumption. Egyptian papyri preserve what is said to be a 
Pharaoh's advice to his son. The message reads: "Hold thyself apart 
from those subordinate to (thee), lest that should happen to whose 
terrors no attention has been given. Approach them not in thy 
loneliness. Fill not thy heart with a brother, nor know a friend. . . . 
(even) when thou sleepest, guard thy heart thyself, because 
no man has adherents on the day of distress." x 

The Arthashastra specifies the dangers which surround the ruler, 
and it discusses the many means by which they can be averted. His 
residence must be made safe. Measures must be taken against poi- 
soning. 2 All members of his entourage must be watched and con- 
trolled. The king must spy on his prime minister. 3 He must beware 
of his close friends,* of his wives, 5 of his brothers, 8 and most particu- 
larly of his heir apparent. According to an authority frequently 
quoted in the classic of Indian despotism, "Princes, like crabs, have 
a notorious tendency of eating up their begetter." 7 To prevent this 
from happening, the manual lists numerous ways by which a ruler 
can protect himself against his son. 8 

b. The Official: Eternal Suspicion 

Nor does the official live securely. "Self-protection shall be the 
first and constant thought of a wise man; for the life of a man un- 
der the service of a king is aptly compared to life in fire; whereas 
fire burns a part or the whole of the body, if at all, the king has the 
power either to destroy or to advance the whole family." 9 

A Persian variant stresses particularly the danger that lurks be- 
hind seeming bureaucratic safety and success: "Should [the ruler] 
at any time pretend to you that you are completely secure with him, 
begin from that moment to feel insecure; if you are being fattened 


by someone, you may expect very quickly to be slaughtered by 
him." 10 

And the need for eternal suspicion is by no means confined to 
those occupying the top of the bureaucratic pyramid. In traditional 
China, as in other hydraulic civilizations, "high officials cannot but 
be jealous of those below them, for it is from that quarter that their 
rivals are to be dreaded. The lower officials, on the other hand, are 
not less suspicious of those above them, for it is from that quarter 
that their removal may be at any moment effected." 


c. The Commoner: the Fear of Being Trapped by Involvement 

The commoner is confronted with problems of a very different 
kind. He is not worried by the pitfalls inherent in autocratic or 
bureaucratic power, but by the threat which this power presents to 
all subjects. A regime that proceeds unchecked in the fields of taxa- 
tion, corvee, and jurisprudence is capable of involving the com- 
moners in endless predicaments. And caution teaches them to avoid 
any unnecessary contacts with their government. 

Smith ascribes the mutual distrust that, according to him, prevails 
in traditional China to the people's fear of getting involved. 12 In the 
Arabian Nights, a corpse is shoved from door to door, because each 
house owner is convinced that the authorities will hold him respon- 
sible for the death of the unknown man. The frequently observed 
reluctance to help a drowning stranger is caused by similar reason- 
ing: If I fail to rescue the poor devil, how shall I prove to the au- 
thorities that I did not plan his submersion? 

Those who walk away when they can be of help are neither dif- 
ferent from nor worse than other human beings. But their behavior 
makes it clear that voluntary participation in public matters, which 
is encouraged in an open society, is extremely risky under conditions 
of total power. The fear of getting involved with an uncontrollable 
and unpredictable government confines the prudent subject to the 
narrow realm of his personal and professional affairs. This fear sepa- 
rates him effectively from other members of the wider community 
to which he also belongs. 

2. The Alienation Potential of Total Power 

Of course, separation is not necessarily alienation: an artisan whose 
forebears left their rural community may consider himself different 
from the inhabitants of his home village. Or an intellectual may feel 
himself out of tune with his co-nationals, or in times of crisis he 
may completely reject a social order that apparently has no use for 

CHAPTER 5, E 157 

him. In such situations he may know loneliness. But as long as he 
can join with others of like mind, his alienation from society will 
be only partial. 

And this partial alienation differs profoundly from total aliena- 
tion. Only when a person believes he is deserted by all his fellows 
and when he is unable to see himself as an autonomous and inner- 
directed entity, only then can he be said to experience total aliena- 
tion. Under the terror of the semimanagerial agrarian apparatus 
state he may know total loneliness without total alienation. Under 
the terror of the modern total managerial apparatus state he may 
suffer total alienation. Persistent isolation and brainwashing may 
bring him to the point where he no longer realizes he is being de- 

3. Every-day Adjustments 

There were many lonely people among the free men of classical 
Greece; a and there are many lonely people in the democratic coun- 
tries of today. But these free individuals are lonely in the main be- 
cause they are neglected and not because they are threatened by a 
power that, whenever it wants to, can reduce human dignity to 
nothingness. A neglected person can maintain associations of some 
kind with a few relatives or friends; and he may overcome his pas- 
sive and partial alienation by widening his associations or by estab- 
lishing new ways of belonging. 

The person who lives under conditions of total power is not so 
privileged. Unable to counteract these conditions, he can take refuge 
only in alert resignation. Eager to avoid the worst, he must always be 
prepared to face it. Resignation has been an attitude of many free 
individuals at different times and in different segments of open and 
semi-open societies. But prior to the rise of the industrial apparatus 
state it was a predominant attitude mainly within the realm of Orien- 
tal despotism. Significantly, stoicism arose in antiquity when the bal- 
anced society of classical Greece gave way to the Hellenistic system 
of total power initiated by Alexander. 

4. Total Loneliness in the Hour of Doom 

The hour of doom realizes what every-day life foreshadows. The 
methods of final destruction operate in one way in a democratically 
balanced world and in another under the rule of total power. 
The free citizen of an open society may fear severe punishment 

a. The tragic and permanent alienation of the slave is too obvious to need elabora- 


/ he core, the margin, and the submargin 
of hydraulic societies 


1. Some Basic Results 

Our inquiry has led to several basic conclusions. First, the institu- 
tional order, hydraulic society, cannot be explained by reference to 
geographical, technological, and economic factors alone. While re- 
sponse to the natural setting is a key feature, it plays a formative 
hydraulic role only under very specific cultural conditions. And it 
involves organizational rather than technological changes. Second, 
some features of hydraulic society appear also in other agrarian 
orders. But hydraulic society is specific in the quality and weight of 
two of its features (hydraulic organization and agrohydraulic des- 
potism). And it is their effective combination that brings into being 
an operational whole, a "going concern" which is able to perpetuate 
itself over millennia. The historian of human freedom must face 
this fundamental empirical fact: among the world's higher pre- 
industrial civilizations, hydraulic society, the most despotic of them, 
has outlasted all others. 

2. Three Problems Deserving Further Investigation 

Why does hydraulic society show such persistence? Is it because of 
its state-managed system of hydraulic agriculture? An upholder of 
the economic interpretation of history will believe this; indeed Marx 
himself argued so. 

But it is significant that Marx and Engels viewed the Tsarist 
government of post-Mongol Russia as Orientally despotic, 1 although 
both certainly knew that Russian agriculture was not hydraulic. 
The difficulty from the standpoint of the economic determinist 
is manifest; and it is increased when we realize that, beside Tsarist 
Russia, certain other agrodespotic states fulfilled the vital organiza- 



tional and acquisitive functions of hydraulic society without main- 
taining a hydraulic economy proper. The capacity of these regimes to 
perpetuate themselves successfully suggests a decisive developmental 
role for the organizational and power features of the agromanagerial 

Obviously the issue is highly important, not only theoretically and 
for the past, but politically and for the present. It is for this reason 
that in this chapter we shall examine the peculiarities and the inter- 
relation of the core and the margin of hydraulic society. In the 
chapters immediately following we shall analyze two other aspects 
of the matter: the power-determined character of private property 
and class rule in the hydraulic world. 

3. Problems of Hydraulic Density 

How hydraulic was hydraulic society? Obviously there are areas of 
maximum hydraulic density and others which, although they are 
hydraulically less dense, may still be considered hydraulic societies 
proper. What is the institutional pattern of the margin of hydraulic 
society? And at what point does this margin lose its societal identity? 
Is there an institutional divide beyond which features of hydrau- 
lic society occur only sporadically in a submarginal form? 

Assuming that such shades of institutional density exist, are they 
static and permanent? Or did hydraulic civilizations shift from the 
margin to the submargin and vice versa? With these questions in 
mind we shall now discuss the core areas, the margin, and the sub- 
marginal zones of the hydraulic world. 


The institutional quality of a hydraulic area varies in accordance 
with its spatial cohesiveness and the economic and political weight of 
its hydraulic system. It may be modified further by the relative 
significance of the second major element of hydraulic operation: 
flood control. 

1. How Continuous Is the Hydraulic System of a 
Given Hydraulic Area? 

The spatial (and organizational) cohesiveness of a given hydraulic 
economy is primarily determined by the continuous or discontinuous 
form of its water supply. A hydraulic commonwealth is apt to create 
a single more or less continuous system of irrigation and flood 
control in a landscape that contains only one major accessible source 
of humidity. Such a development frequently occurs in oasis-like 

CHAPTER 6, B 163 

regions crossed by a river that gathers the bulk of its water in a more 
humid hilly or mountainous hinterland. The river-valley states of 
ancient coastal Peru maintained continuous hydraulic systems. In 
the Old World, Sindh and the Nile Valley civilization of Egypt are 
classical variants of the same pattern. 

If an arid landscape includes several not too widely separated 
rivers, the canals leading from them may form a relatively continuous 
hydraulic network. However, few arid regions are so privileged. 
Lower Mesopotamia is more the exception than the rule. 

In most cases the rivers of a potentially hydraulic landscape lie 
too far apart to permit interlocking through connecting canals. Con- 
sequently a hydraulic commonwealth covering a multi-river area 
generally maintains a discontinuous system of embankments and 
canals. Individuals depending on a limited and single water supply 
may reproduce a limited tribal or national culture for a long period 
of time. This happened in the Rio Grande area and, on a much more 
impressive scale, in Pharaonic Egypt. But the self-perpetuating hy- 
draulic tribes played an insignificant part on the stage of human 
history; and even such national complexes as Egypt eventually out- 
grew their early political isolation. The great majority of all his- 
torically conspicuous hydraulic nations and empires include regions 
which depend on a continuous hydraulic unit; yet, taken as a whole, 
the hydraulic system of these larger political units have a definitely 
discontinuous form. 

2. How Great Is the Economic and Political 
Weight of a Given Hydraulic Economy? 

Since most of the larger hydraulic civilizations maintain discon- 
tinuous hydraulic systems, lack of cohesiveness obviously is no reli- 
able index for establishing hydraulic density. The economic and 
political weights of a discontinuous hydraulic system must be estab- 
lished by other means. 

In arid areas a discontinuous hydraulic system occurs occasionally; 
in semi-arid areas it is virtually the rule, at least for societies that 
have outgrown their most primitive beginnings. As indicated above, 
the semi-arid areas which have given rise to hydraulic developments 
are numerous and large; and within them the relation between 
hydraulic agriculture and nonhydraulic (small-scale irrigation and 
rainfall) farming varies enormously. 

Three major shades of this relation can be distinguished: 
1) The hydraulically cultivated land may comprise more than half 
of all arable land. Since hydraulic agriculture tends to produce 


yields that, by and large, are as high as those produced by small- 
scale irrigation and definitely higher than the average yields of the 
rainfall farmers, a hydraulic agriculture which covers more than 50 
per cent of all arable land may be said to be in a position of absolute 
economic superiority. 

This condition is found most frequently in arid regions; and 
frequently, although not necessarily, it is found together with a 
continuous hydraulic system. In most Rio Grande Pueblos the bulk 
of all land is irrigated; and the bulk of irrigation water is drawn from 
communally operated irrigation ditches. In Egypt, from the dawn of 
history, the great majority of all fields was irrigated either by inunda- 
tion or through canals. 1 In the delta a meager crop can be grown 
by methods of rainfall farming; ° and throughout the country, wells 
can be used to water vegetables, gardens, and orchards. 2 But as in 
the case of the Rio Grande Pueblos, these supplementary forms of 
cultivation do not challenge the overwhelming economic superiority 
of the hydraulic economy. 

2) The hydraulically cultivated land, even when it comprises less 
than half the country's arable acreage, may nevertheless yield more 
than all other arable land. In this case, hydraulic agriculture may be 
said to hold a position of relative economic superiority. On the eve 
of China's unification the state of Ch'in enormously strengthened its 
agrarian heartlands (in present Shensi) by constructing the Cheng 
Kuo irrigation works; and this action made Ch'in richer and more 
powerful than any other territorial state. In the subsequent period, 
the whole area of what had been Ch'in h comprised about one-third 
of the empire's area, but, according to Pan Ku, it accounted for 60 
per cent of its wealth. 8 Ssu-ma Ch'ien considered the former Ch'in 
territory "ten times as rich as [the rest of] the empire." * Neither of 
these statements can be verified, and they certainly should not be 
pressed. Yet they illustrate what we mean by the relative economic 
superiority of a vigorous hydraulic system of agriculture. 

3) The hydraulically cultivated land, even if it is inferior both in 
acreage and yield to the remaining arable land, may nevertheless be 
sufficient to stimulate despotic patterns of corvee labor and govern- 
ment. In this case the larger, nonhydraulic area essentially produces 
food, whereas the smaller, hydraulic area, in addition to producing 

a. After mentioning the cultivation of barley in the Nile delta as one of the examples 
of rainfall agriculture close to the minimum limit, the Agricultural Yearbook of 1941 
concludes: "Production year after year with these small amounts of moisture is possible 
only where the distribution of rainfall during the year and other climatic conditions 
are favorable and where the moisture falling in two or more years is stored for one 
crop" (CM: 322). 

b. In addition to the Cheng Kuo complex, this included among other regions the 
classical irrigation plain of Szechwan. 

CHAPTER 6, B 165 

food, produces power, and it produces power that is sufficiently 
strong and sufficiently despotic to control both sectors of the agrarian 

This evidently happened in numerous semi-arid regions that were 
suitable — in key areas — for hydraulic operations. During the forma- 
tive period of many great hydraulic civilizations despotic power 
probably arose under exactly such conditions; and the pattern has 
been perpetuated in historic times. Assyria and Mexico applied 
methods of mass control that were imperative only in relatively small 
hydraulic regions to large areas of small-scale irrigation and rainfall 
farming. Under these conditions the hydraulic economy, though pre- 
dominant neither in acreage nor yield, nevertheless occupied a posi- 
tion of organizational and political superiority. 

3. How Strong Is the Second Major Element of 
Hydraulic Operation: Flood Control? 

Where the hydraulic system prevails economically, the relative 
strength of protective (as compared with productive) water works 
is of little concern. An elaborate hydraulic agriculture involves an 
elaborate bureaucratic development; and the despotic regime is thus 
conveniently bulwarked. 

Things are different when the hydraulic system, although suffi- 
cient to establish political supremacy, involves only modest bureau- 
cratic developments. To be sure, the maintenance of large installa- 
tions for flood control always necessitates comprehensive operations 
of mobilization and on-the-spot direction; and it also heightens the 
quasimilitary authority of the managerial government in situations of 
absolute or relative economic hegemony. But the protective factor 
becomes particularly important when economic hegemony is lack- 
ing. The fight against large and disastrous floods tends to expand 
government-directed mass mobilization further than would produc- 
tive hydraulic action alone. And the disciplinary measures involved 
in protective enterprises do much to cement the power of a govern- 
ment that derives only a limited managerial authority from its 
agromanagerial achievements. In the lake area of ancient Mexico 
the struggle against periodic and devastating floods probably re 
quired much larger corvee teams than did the regional irrigation 
works. The significance of this fact for the aggrandizement of govern- 
ment power can be easily imagined. 

4. Compact and Loose Hydraulic Societies 

Our argument does not exhaust all morphological possibilities. But 
it establishes one point beyond doubt: The core areas of the hydrau- 


lie world manifest at least two major types of hydraulic density. 
Some are hydraulically compact, whereas others are hydraulically 
loose. 5 A hydraulic society may be considered "compact" when its 
hydraulic agriculture occupies a position of absolute or relative 
economic hegemony. It may be considered "loose" when its hydraulic 
agriculture, while lacking economic superiority, is sufficient to assure 
its leaders absolute organizational and political hegemony. 

This primary division may be supplemented by some important 
secondary divisions. A hydraulic society, whose hydraulic agriculture 
is economically dominant and spatially continuous, is an extreme 
variant of the compact pattern (C i). A hydraulic society whose 
hydraulic agriculture is economically dominant but discontinuous is 
a less extreme variant of this same pattern (C 2). Distinction between 
absolute (a) and relative (r) economic hegemony enables us to carry 
the differentiation still further (Ca 1 and Cr 1, Ca 2 and Cr 2). 

A loose hydraulic society may include among its installations large 
units which are compact within their immediate locale or which go 
beyond the borders of a single region. The relatively great hydraulic 
weight of this pattern may be indicated by the symbol "L 1." A 
loose hydraulic society whose largest hydraulic units fail to achieve 
economic hegemony even regionally represents the lowest hydraulic 
density type (L 2). Another differentiating factor, the relatively 
strong development of protective hydraulic works, may be indicated 
whenever this seems desirable by the formula "-f- prot." 

A few examples indicate, on a tribal or national scale, the four 
main categories of hydraulic density: 

Compact 1: Most Rio Grande Pueblos, the small city states of 
ancient coastal Peru, Pharaonic Egypt. 

Compact 2: The city states of ancient Lower Mesopotamia, prob- 
ably the state of Ch'in on the eve of the Chinese empire 

Loose 1: The Chagga tribes, ancient Assyria, the old Chinese state 
of Ch'i (L 1 + prot.), and perhaps Ch'u. 

Loose 2: Tribal civilizations: The Suk of East Africa, the Zuni of 
New Mexico. State centered civilizations: indigenous Hawaii, 
many territorial states of ancient Mexico (L 2 + prot.). 

5. The Great Agromanagerial Empires — Usually 
Loose Hydraulic Societies 

Dominion of one city state over a number of other city states leads 
to the establishment of rudimentary empires. Conformations of this 
kind arose in ancient Lower Mesopotamia, on the coast of ancient 
Peru, in Chou China, and in Buddhist India. 

CHAPTER 6, B 167 

In the first two cases the components were of the compact hydrau- 
lic type; and the quasi-imperial units were also hydraulically com- 
pact. Usually, however, military and political expansion resulted in 
the creation of larger and less homogeneous conformations. The 
great hydraulic empires tended to include territorial and national 
units of different hydraulic densities. They formed loose hydraulic 
societies, which frequently included compact hydraulic subareas. 
The Babylonian and Assyrian empires, China during the periods of 
unification, the great empires of India, Achaemenian Persia at the 
height of its expansion, the Arab caliphate, Ottoman Turkey, the 
Inca empire, and the federation of Aztec Mexico — all were hydrau- 
lic societies, and all, perhaps with the exception of Mexico, belonged 
to the category L 1 . 

The hydraulic glands of the great agromanagerial empires have 
been accorded little systematic attention. A morphological study of 
the hydraulic order of traditional China reveals many density patterns 
and significant super-regional arrangements. Mez' thoughtful anal- 
ysis of Abbassid power indicates the number and variety of the 
great hydraulic areas that for shorter or longer periods lay within 
the jurisdiction of the Baghdad caliphate: Egypt, South Arabia, 
Babylonia, Persia (northeast and south Transoxania and Afghani- 
stan). 6 All these areas posed "great irrigation problems," 7 and the 
Arab sources note both the technological means and the numerous 
personnel required to solve them. 8 

6. Degrees of Hydraulic Density and Degrees of 
Bureaucratic Density 

a. The Principle 

The bureaucratic density of an agromanagerial society varies with 
its hydraulic density. This correlation is affected by such factors as 
the institutional weight of large nonhydraulic constructions (the 
Zuni Pueblos, the territorial states of Chou China, the Roman em- 
pire) and the dimensions of communicational and/or military or- 
ganizations (Assyria, the state of Ch'in, Aztec Mexico). But such 
factors modify rather than negate the basic hydraulic-bureaucratic 
relation. Pharaonic Egypt was highly bureaucratized long before it 
developed a comprehensive military officialdom. And while both the 
Incas and the Aztecs maintained strong military organizations, there 

c. For a discussion of the varying territorial dimensions and character, as well as 
the interarea relations, in the "loose" hydraulic order of traditional China, see 
Wittfogel, 1931: 252-72. 


can be little doubt that the former had a more comprehensive 
managerial bureaucracy than the latter. 

On the acquisitive level correlations also vary. To be sure, an 
agrarian despotism, no matter what its hydraulic density pattern, 
insists upon its right to tax universally. Yet the way in which this 
right is exercised differs significantly. Although a loose hydraulic 
society with a strong government may be able to gather in a larger 
percentage of the estimated revenue than a compact hydraulic society 
with a weak government, other conditions being equal, the more 
comprehensive bureaucracy of an intensively managerial state is 
better equipped over time to handle the business of taxation than is 
the less comprehensive bureaucracy of a less intensively managerial 

The collecting of the rural surplus was more centralized in Inca 
Peru than in Aztec Mexico, where local affairs were handled not by 
representatives of the government but by heads of the local calpulli. 
In the compact hydraulic societies of the ancient Near East the bulk 
of the revenue seems to have been gathered by government func- 
tionaries, although intermediaries are known to have been used in 
certain periods in Pharaonic Egypt. 9 Under Greek and Roman in- 
fluence respectively, tax farming appeared in the Hellenistic and 
Roman Near East; 10 but the absolutist regimes soon asserted their 
power, first by modifying the system of tax farming and later by 
reducing it to insignificance. 11 State-appointed (liturgical) tax col- 
lectors, mostly wealthy townsmen, supplemented the fiscal bureauc- 
racy; and big (bureaucratic) landowners fulfilled a similar function 
with more advantage as well as less danger to themselves. 12 Thus 
the hydraulically loose Roman empire discarded the independent 
tax farmers of ancient Greece and republican Rome without revert- 
ing to the old Egyptian and Babylonian ways of directly and bureau- 
cratically collecting the revenue. 

This step was taken by the Arab masters of the Near East, whose 
power was rooted in such hydraulic centers as Damascus, Cairo, and 
Baghdad. Under the Umayyads the bureaucratic fiscal system pre- 
vailed; and the tax farmers, whom the Abbassid government began 
to employ, were still closely integrated in the bureaucratic order. In 
Mesopotamia they were part of the officialdom. 13 In China some local 
tax collectors were not members of the regular officialdom; 14 but 
bureaucratic methods of tax collection seem to have prevailed 
throughout the ages. 

CHAPTER 6, B 169 

b. Changing Bureaucratic Density of a Hydraulic Territory 

The inclusion of incipiently hydraulic or nonhydraulic territories 
in a loose hydraulic society is usually followed by the development 
of a bureaucratic network in these territories. This is what happened 
when the ancient centers of Chinese culture conquered certain "bar- 
barian" regions in Central and South China. 

The inclusion of a compact hydraulic territory in a hydraulically 
loose empire tends to have the opposite effect. The rulers, who are 
accustomed to operate with a less compact officialdom, may also 
reduce the bureaucratic apparatus of the hydraulically compact 
area. This is what happened when the Nile Valley became part of 
the Roman empire. 

7. Hydraulically Concerned and Hydraulically 
Unconcerned Masters of Hydraulic Society 

A second factor that may change the bureaucratic density of a 
hydraulic society is the rulers' concern (or lack of concern) for hy- 
draulic management. As discussed previously, a hydraulic society may 
sink to a low rationality level if it is ruled by conquerors who take 
little interest in managerial agriculture or if its indigenous masters 
slacken their productive efforts. The conquerors' lack of hydraulic 
concern is usually a consequence of their nonhydraulic background. 
Internal decay may be due to a reduction in government revenue 
resulting from the excessive growth of proprietary forces or from the 
degeneration of a ruling group that reveled in the luxury of total 

The spatial relation between the main areas of political power 
and hydraulic economy also plays a part. Rulers may establish their 
capital close to the major regions of agricultural wealth and sur- 
plus; or they may establish it at a considerable distance from these 
regions. Defense is often given as the reason for the latter decision, 
and at times it may indeed be the whole reason. Often, however, the 
rulers — particularly conqueror-rulers — preferred to set up their capi- 
tals in a nonhydraulic frontier, because they had a stronger affinity 
to the periphery than to the core areas of the hydraulic world. 

In China the centers of political direction and hydraulic economy 
coincided more or less until the first millennium a.d., when the 
growing fertility of the Yangtze area conflicted with the defense 
needs of the vital northern border zone. From then on, the seat of 
the central government shifted back and forth; but the northern 
region never ceased to be hydraulic to some extent, and the northern 


capitals were ingeniously and hydraulically connected with the main 
rice areas of Central China through the Grand Canal. 

In India the great northern plain, which was the main area of 
hydraulic agriculture, was also the logical place for the political 
metropolis; and the Muslim masters of India, like their Hindu 
predecessors, established their capitals there. But they exhibited less 
hydraulic concern than had the previous indigenous rulers. Although 
they were not lacking in managerial interest, and although they 
created and maintained large irrigation works, they never fully 
restored the grandiose hydraulic economy that appears to have 
flourished in the Maurya empire. The role they assigned to local 
"chiefs" and tax farmers reflects the relatively low bureaucratic 
density of Muslim India. 

The later Roman emperors responded to the lure of the East. 
Yet they established their new capital, not in one of the great classical 
areas of hydraulic agriculture (Egypt, Syria, or Mesopotamia) but at 
the Hellespont, the classical divide between the Orient and the non- 
hydraulic West. And despite the fact that long acquaintance with 
managerial despotism stimulated them to plan and build on a large 
scale, they were content to administer their hydraulic possessions 
from afar. Immensely bold in the creation of nonhydraulic con- 
structions (highways and frontier walls), they exhibited much less 
initiative in the agromanagerial sphere. While by no means lacking 
in hydraulic concern, they aimed at gathering as large a rural 
revenue as possible with as small a bureaucracy as possible. Rational 
rulers though they were, they did not realize the rationality maxi- 
mum of the hydraulic world they controlled. 

The Romans, who made Constantinople the capital of their em- 
pire, had behind them five hundred years of practical experience 
with the Hellenistic version of hydraulic statecraft. The Turks, who 
had conquered Adrianople in 1361, Constantinople in 1453, Egypt 
in 1517, and Mesopotamia in 1534, were not unacquainted with 
higher agrarian civilizations of the hydraulic type either; as a 
matter of fact, they had lived at the edge of the hydraulic world since 
the dawn of history. But perhaps because of their pastoral back- 
ground they were less interested in the promotion of agriculture 18 
than in military enterprises; and they preferred extending the non- 
hydraulic margin to intensifying the hydraulic core. True, the great 
irrigation works of Mesopotamia lay in ruins when the Turks came; 
but the history of China and India shows that hydraulic effort can 
restore quickly what antihydraulic action has destroyed. The Turks 
did not break with agromanagerial tradition in Egypt or Syria; but 
they furthered no significant reconstruction work in Iraq. Speaking 

CHAPTER 6, B 171 

generally, they displayed no effective zest for hydraulic develop- 
ment. 16 As Orientally despotic organizers of war, peace, and fiscal 
exploitation, they were extraordinarily successful; and in some few 
major administrative centers they employed many officials. Being 
managerially unconcerned, however, they governed their far-flung 
empire with a relatively small professional bureaucracy. 

8. Periods of Agromanagerial Adjustment, 
Degeneration, and Restoration 

Of course, the economic ethos (the Wirtschaftsgesinnung) of a 
ruling group is not unchangeable. Great differences in cultural and 
social assimilation notwithstanding, this is true also for pastoral 

The tribal conquerors of China were usually willing to uphold 
the indigenous tradition in certain spheres of nonhydraulic con- 
struction and management; and many of them became at least super- 
ficially aware of the importance of irrigation agriculture. Perhaps 
none of the northern conquerors equaled the active hydraulic con- 
cern of the Manchus, who had practiced irrigation in their homeland 
prior to their conquest of China. 17 In the Near East the Umayyads, 
who consolidated a conquest regime established by the first followers 
of the Prophet, also showed extraordinary hydraulic concern. 18 

Pastoral and semipastoral conquerors who develop an interest in 
hydraulic matters do so, as a rule, not during the first period of their 
dominion but later; and often they grow managerially lazy and 
negligent before their rationality potential has been exhausted. In- 
digenous rulers, on the other hand, frequently show the greatest 
hydraulic concern during the earlier periods of their regime, tending 
to grow managerially less insistent when their power is consolidated. 
In either case, decay may be retarded by challenging external cir- 
cumstances; or it may be accelerated by the expansion of large 
proprietary forces, whose representatives arrogate to themselves an 
increasing part of the national surplus. 1 * When one segment of the 
despotic elite (primarily the court and clusters of officials close 
to it) succumbs to the corrupting influence of total power, another 
segment (other members of the officialdom and their relatives and 
friends among the bureaucratic "gentry") may seize power. As the 
result of this process, excessively irrational features may be eliminated 
in a "cathartic" and "regenerative" revolution. 

d. For an attempt to explain the great agrarian and political crises in Chinese 
society by means of this and other social factors see Wittfogel, 1927: 322 ft., 328 ff.; 
ibid., 1935: 53. Cf. Wittfogel and Feng, 1949: 377. For an analysis of agrarian crises 
as a general feature of Oriental society see Wittfogel, 1938: 109 ff. 


A development of this type does not change the traditional hy- 
draulic and despotic order; it merely restores its vitality. The first 
rulers of many Egyptian, Babylonian, Chinese, Indian, Persian, 
Islamic, and Mexican dynasties have been praised for their vigor 
and efficiency. Regenerative upsurges may also occur during a later 
phase of a dynastic reign; and then, as during the formative period, 
serious attempts may be made at effective hydraulic management 
and rational fiscal administration. In both cases the more far- 
sighted and less compromised elements within the ruling bureauc- 
racy demonstrate that they can run the country in a more effective 
way than their self-indulgent and "corrupt" rivals. 

9. The Staying Power of Deteriorated Agro- 
managerial hydraulic societies 

The dominant myths of Oriental despotism ascribe regenerative 
achievements to almost every founder of a new dynasty; but an un- 
biased evaluation of the evidence leads to less flattering conclusions. 
Under conditions that permit no independent criticism or political 
pressure, the immediate benefits of total power have a much greater 
appeal to the masters of the absolutist apparatus than do the 
potential fruits of rational— albeit, selfishly rational — managerial 
effort. Self-indulgence is, therefore, a more typical motive for be- 
havior than the desire to maintain the rulers' rationality optimum. 

And this is true not only for most later sovereigns but also for 
many a dynasty's founding father. Such persons, however vigorous, 
are often more sensitive to the political weaknesses of the old regime 
than to the managerial possibilities of the new. Having won over the 
bulk of the military and civil officials, they readily correct the most 
glaring abuses in taxation, forced labor, or jurisdiction, and they 
make the most urgent constructional and agromanagerial improve- 
ments; but they have neither the vision nor the personnel to raise 
the hydraulic government to a conspicuously higher level of hydrau- 
lic and fiscal management. In the many dynastic changes that charac- 
terize the history of agromanagerial civilizations, thorough re- 
generative upsurges are probably more the exception than the rule. 

Of course, a stoppage of all hydraulic operations would paralyze 
agricultural life, and this not only in areas of full aridity but in 
many semi-arid regions as well. Consequently, even a hydraulically 
unconcerned Oriental government will devote some effort to its 
managerial duties. It has to carry on somehow, even if it must depend 
largely and not too rationally on local groups. During the last phase 
of Byzantine rule over Egypt, influential landlords, most of whom 

CHAPTER 6, C 173 

had bureaucratic connections, 19 are said to have maintained the 
dikes and canals in many localities. 20 To what extent governmental 
hydraulic action was reduced by this arrangement is hard to decide. 
Even during this critical period, however, Egypt's irrigation economy 
was sufficiently continuous and sufficiently effective to feed the people 
and to furnish a huge revenue. Somehow it succeeded in perpetuating 
itself. When the Arabs appeared in 639, they found in the Nile 
Valley a population of about seven millions, 6 that is, about as many 
persons as had lived there under Ptolemaic rule. 


In arid or semi-arid landscapes sedentary agrarian civilizations can 
persist permanently and prosperously only on the basis of a hydraulic 
economy. Along the moderately humid periphery of the arid and 
semi-arid world agrarian life is not so conditioned. Here Oriental 
despotism may prevail with little or no dependence upon hydraulic 

1. Varying Operational and Bureaucratic Density 
Patterns in Marginal Areas of the Hydraulic 

In the hydraulic core areas degrees of hydraulic density provide a 
crucial means for distinguishing degrees of institutional density. In 
the margins, however, this criterion loses its significance. Instead, 
degrees of bureaucratic density are best determined by an approach 
that evaluates the relative development of absolutist methods in the 
spheres of construction (mostly nonhydraulic), organization, and 

Comparison between the states of Middle Byzantium and post- 
Mongol Russia reveal significant differences. Byzantium maintained 
considerable hydraulic installations, in the main for providing drink- 
ing water; * and these have no parallel in Muscovite Russia. Nor 
did the Muscovite Russians engage in comprehensive nonhydraulic 
constructions as did the Byzantines. The founders of Eastern Rome 
reshaped the earlier network of roads; * and their highways were 
the foundation of the Byzantine system of communications, 2 which 
in a limited way continued in use even under the Turks. 3 

e. For the beginning of the Arab era see Johnson and West, 1949: 263 (6,000,000, 
plus children and old people); cf. Munier, 1932: 84. For Ptolemaic Egypt see 
Diodorus, I, sec. 31 (7,000,000); cf. Josephus, JW 2.16 (7,500,000); Wilcken, 1899, I: 489 ff. 

a. Br^hier, 1950: 90 ff. For a description of some of these works see Ritter, 1858: 
155, 160, 167, 202, 346, 378, 406, 496, 547. Most of the local and regional hydraulic 
works that existed under the Turks probably go back to the Byzantines. 


The Byzantines also made enormous building efforts for purposes 
of defense. They protected their borders by a great chain of fortifica- 
tions; and here, as in the sphere of communications, corvee labor 
was mobilized for the task.* After the victory of the Seljuq Turks 
at Manzikert (in 1071), the absolutist state still functioned; and the 
road corvee was still levied in the 12th century; 5 but the vigor of the 
early days was gone. The great military road, which in the preceding 
years had had its periods of decay and reconstruction, appears to 
have been properly maintained only "until the eleventh century." 6 

When the Mongols established their rule over Russia, they did 
not construct massive roads, nor did they erect frontier walls or 
chains of border fortresses. They were satisfied to establish organiza- 
tional and acquisitive methods of total control. It is in these two 
last fields of action that Byzantium and absolutist Russia, although 
not identical, were similar. 

The Byzantines kept account of their country's wealth in elaborate 
cadasters. 7 They monopolized quick communication and intelligence 
by means of the state post. 8 They closely controlled the major sectors 
of handicraft and commerce, again until the 1 1 th century. 9 And they 
maintained armies whose orderly integration contrasted strikingly 
with the amorphous hosts of feudal Europe. 10 

All these features have parallels in Muscovite Russia. The mature 
Muscovite state registered the mass of its population for fiscal and 
military purposes; " it operated an elaborate "postal" (relay) sys- 
tem; 12 it occupied a key position in the country's trade; 13 and it 
despotically drafted and directed its fighting men." 

During the earlier periods of both absolutist regimes office land 
was assigned to persons serving the state. In Byzantium this system 
emerged on the eve of the Arab conquest in a time of turmoil and 
invasion and as a means of strengthening defense against the Persian 
attack. Rooted in earlier Roman institutions 14 and set in its classi- 
cal form by Heraclius I (610-641), it continued patterns that had 
existed in the ancient Orient from the days of Sumer and Babylon 
and that prevailed also in contemporary Persia. 18 Under the system 
of themes, each Byzantine soldier received a farm which, like his 
service, was hereditary and indivisible. 16 

This plebeian version of an absolutist office land system lasted 
until the 1 1 th century. Then, after the catastrophic defeat at Manzi- 

b. For the principle see Herberstein, NR, I: 95 ff.; for its full development, Staden, 
1930: 58; cf. Kluchevsky, HR, II: 48, in, 115. As will be shown below, all these in- 
stitutions existed before Ivan III (1462-1505), during whose reign the Tatar Yoke 

CHAPTER 6, C l^tj 

kert, the state placed at the -center of its reorganized military (and 
office land) system the big landowners, who, with the development 
of a heavy cavalry, were more useful militarily than the themes 

Hand in hand with this transformation went the transformation of 
the acquisitive order. From the 7th to the 1 1 th century the govern- 
ment collected the bulk of its revenue through its officials. The 
themes soldiers, who lived essentially off their service land, presented 
no major fiscal problem.'* The holders of the pronoia, the larger land 
units that constituted the core of the later office land system, pro- 
vided a certain number of heavily armed soldiers and collected taxes 
from the peasants of the pronoia. 17 Together with the newly estab- 
lished tax farmers, 18 the pronoetes formed a group of semi-official 
tax collectors, who were less directly controlled by the state than 
were the members of the regular fiscal bureaucracy. 

The corresponding Russian development has certain distinct 
features. The Muscovite holders of office land, the pomeshchiki, in- 
sofar as they rendered military service, were from the beginning and 
in the main heavily armed horsemen, and because of the greater 
burden of their equipment they were usually assigned estates larger 
than a peasant farmstead. Within their pomestye they collected taxes 
from their peasants. Consequently their government, like the govern- 
ment of later Byzantium, gathered only a part of its revenue through 
professional fiscal officials. 

Both regimes employed despotic methods of government in the 
organizational and acquisitive fields. In the constructional field such 
methods were used to a major degree only by Byzantium, and there 
essentially during the middle period (until the 11th century). The 
shrinking range of constructional operations in post-Manzikert 
Byzantium was interestingly paralleled by the shrinking range of its 
fiscal bureaucracy. In Muscovite Russia constructional activities 
were irrelevant from the start; and the fiscal system was, also from 
the start, characterized by a large nonbureaucratic sector. 

Thus a positive correlation between operational and bureaucratic 
density can be formulated for the margin as well as for the core 
areas of hydraulic society. This correlation may be influenced by 
other factors, and strongly so. But experience bears out what the- 
oretical considerations suggest: Other conditions being equal, the 

c. Cf. Ostrogorsky, 1940: 262. Ostrogorsky describes the military difference between 
the two groups, which I correlate here with the two types of office land. 

d. Ostrogorsky, 1940: 58. According to the Tactica Leonis 20.71, they seem to have 
paid some minor imposts (ibid.: 48). 


density of the despotic bureaucracy tends to increase or decrease 
with the increase or decrease of its functions. 

2, The Growth of Proprietary Forces 

In Byzantium and pOst-Mongol Russia the state controlled the bulk 
of the land either fiscally or administratively, a large part of it being 
assigned as office land to the soldiers of the themes, the pronoetes, 
or to the pomeshchiki. Socially and economically, the pronoia holders 
were more powerful than the plebian peasant warriors of the themes; 
but they bore a closer resemblance to the Russian pomeshchiki than 
to the feudal lords of Western Europe. Both the pronoetes and 
pomeshchiki delivered part of their rural revenue to the state. Both 
owed absolute obedience to their respective governments. And both 
lacked the decisive capacity of feudal and postfeudal landlordism — 
the capacity to organize independent nationwide political corpora- 
tions (estates, stdnde). 

However, these conditions did not prevail unaltered. They existed 
in Late Byzantium up to 1204, the year in which the completely 
defeated empire was replaced by the Latin Empire; and they under- 
went a great change in the final period of Byzantium, which ended 
in 1453. In Russia they existed up to 1762, the year in which the 
former pomestye land became the private property of its holders. 

In later Byzantium and in post-Muscovite Russia private property 
and enterprise gained considerable strength. In view of this fact 
we may ask first, is such a development typical of agrarian des- 
potisms and second, to what degree was the growth of proprietary 
forces responsible for the societal changes that occurred in Byzantium 
from 1261 to 1453 and in Russia from 1861 to 1917? 

In Byzantium big landownership was an important factor even 
before 1071; but its significance increased greatly when, at the end 
of the 11th and at the beginning of the 12th century, the landlord- 
pronoetes were given additional economic and judicial power. After 
the fall of the Latin Empire, the pronoetes, who formerly had held 
their grants for a limited time only, achieved the "hereditary and 
unrestricted ownership" of their lands. And they also obtained tax 
exemptions far greater than anything that had been customary. 19 The 
corresponding shrinkage in the government revenue was a decisive 
factor in the weakening of the Byzantine empire, which eventually 
was unable to resist the Turks. 

In Tsarist Russia events took a different course. Here industrializa- 
tion made substantial advances in the 18th and particularly in the 
19th century; and this development was closely related to the growth 


of private property, first immobile (land) and ultimately also mobile 

3. The Institutional Staying Power of Marginal 
Oriental Despotism 

But the growth of proprietary forces did not bring about a trans- 
formation in Byzantine society like that achieved in Western Europe. 
Nor did it, prior to 1917, enable the Russian men of property to pre- 
vail over the men of the state apparatus. Why not? Were the bene- 
ficiaries of total power fully aware of the issue involved? And did 
they aim at isolating and crippling the representatives of property? 

It is easy to juxtapose neatly separated camps. The real conditions, 
however, were much more complicated. In Byzantium, in Tsarist 
Russia, and in most other Orientally despotic countries the men of 
the apparatus were frequently also men of property. Consequently 
the conflict between the interests of the absolutist regime and the 
interests of private property and enterprise appear also — and often 
primarily — as a conflict between different members of the same rul- 
ing class or even as a conflict between different interests of individual 
members of this class. Why do such persons — as a group and over 
time — place their bureaucratic above their proprietary interests? 

a. Bureaucratic Interests Favoring the Reproduction of the 
Despotic Order 

The civil or military official of an agrarian despotism is part of a 
bureaucratic hierarchy, which, taken in its entirety, enjoys more 
power, revenue, and prestige than any other group in the society. 
Of course, the post he holds today and the one he hopes to hold 
tomorrow carry with them the risk of total destruction; and he is 
therefore never safe. However, under the shadow of total power the 
man of property is never safe either; and the dangers of his position 
are not outweighed by satisfactions derived from active participation 
in the gambles and privileges of total power. Thus, not even the 
members of the bureaucratic class who hold no office challenge the 
principles of the absolutist regime, which they may rejoin tomorrow. 
And the officiating members of this class, confronted with the- Big 
Conflict, aggressively uphold the privileges of bureaucratic power, 
revenue, and prestige which they are enjoying now. 

Narrow and oversimplified interpretation has obscured the issue 
by formulating it only in terms of the interests of a single person, 
the autocratic ruler. To be sure, the despot is eager to perpetuate 
his absolute power, but, lacking an effective governmental apparatus, 


he cannot achieve this aim. The kings of Medieval Europe found 
absolutist power as sweet as did their Byzantine confreres. But the 
latter succeeded where the former failed, because the integrated 
Byzantine bureaucracy upheld the system of total power that favored 
both the sovereign and the men of the apparatus, whereas the 
enfeoffed vassals of the Western kings safeguarded and reproduced 
their privileges by keeping the king's power fragmented and checked. 
To what extent can the prominence of the army in certain agro- 
managerial countries be taken as a sign of feudal decentralization? 
Military officials are as much men of the state apparatus as are their 
civil opposites; and if the first centuries of the Roman empire 
demonstrate anything, it is exactly this. For it was just when military 
leadership was prominent that Roman absolutism attained its matu- 
rity. The crystallization of despotic power in Muscovite Russia in- 
volved considerable bureaucratic activity; but the overwhelming 
majority of the new serving men wielded the sword and not the pen. 
The fact that in later Byzantium the heads of the military sector of 
the state apparatus figured prominently also as political leaders re- 
flects the increasing pressure of foreign aggression. But it does not 
mean that these individuals served their government in a limited 
and conditional way as members of a baronial and feudal class. 

b. Late Byzantium: Marasmus rather than Creative 

We must remember all this when we try to evaluate the effect of 
big property on the society of later Byzantium. Landed property in- 
creased during the first centuries of the Middle Empire; yet state 
protection of peasant holdings and periodic confiscations of large 
estates 20 notably retarded this development. After 1071, controls 
grew looser, but the state still had a rein on the country's rural 
economy. Contrary to corresponding developments in feudal Europe, 
conversion of the cadaster from a public to a private institution 
"never occurred in the East." 21 And the pronoetes, however they 
may have benefited personally, had to deliver a large part of the 
taxes they collected to the government. 22 

After the interlude of the Latin Empire, the state of Byzantium 
never regained its earlier authority. The landowners were now strong 
enough to withhold a much greater proportion of the national sur- 
plus than they had done previously, but they did not consolidate 
their ranks. Neither the great landowners nor the representatives of 
mobile urban wealth established nationwide corporations: estates. 
Private property became big; but it remained politically unor- 

CHAPTER 6, C 179 

ganized. Contrary to corresponding developments in the West, the 
growth of big private property in Byzantium did not give birth to a 
new society. It succeeded only in weakening and paralyzing the old 

c. The Extraordinary Staying Power of Tsarist Bureaucracy 

After 1204 the Latin Empire temporarily replaced the traditional 
despotic regime. Could it be that the quasifeudal institutions of this 
empire (and of the Western enemies of Constantinople in general) 
influenced the bureaucratic absolutism of Byzantium so seriously 
that it was never able to regain its former superiority? In other words, 
did the rural and urban proprietors succeed in paralyzing the 
Byzantine government in the last centuries only because external 
forces broke the backbone of despotic power? 

In terms of the fundamental issue the experiences of Tsarist 
Russia are eminently instructive. Post-Mongol Russia was invaded 
several times; but prior to the democratic revolution of 1917 the 
absolutist government was never completely broken. Russia's in- 
dustrialization was strongly stimulated by Western developments. 
Foreign money flowed into private (capitalist) enterprises, increasing 
the weight of the proprietary sector. And Western methods and 
ideas notably affected Russian thought and performance. But all 
these external influences did not destroy the absolutist character of 
the state. The relation of the Tsarist bureaucracy to the forces of 
property — and eventually also to labor — continued to be determined 
by conditions that had long been operative in traditional Russian 
society. And this relation was, and remained, a relation of absolute 
bureaucratic superiority. 

The masters of the despotic state apparatus responded to the 
changing historical situation with changing attitudes, but until 1917 
they did not relinquish their total power. When in the early 18th 
century it became obvious that industrialization was vital for the 
country's defense, the Tsarist government was not satisfied with 
supervising and regulating the new industries, as the absolutist 
governments of Western Europe were doing. Instead, it directly 
managed the bulk of the heavy industry and, in addition also, part 
of the light industry/ probably employing for these purposes the 

e. In 1743 the state had some 63,000 male "souls" ascribed to its (Ural) Mountain 
Works and 87,000 "souls" to its potash works (Mavor, 1925, I: 441), plus an un- 
known number of individuals who labored outside of these two main spheres of 
government production, whereas private workshops and factories occupied some 30,000 
(ascribed) male "souls" (ibid.: 493). Under Elizabeth (1741-62) the sector of state- 


majority of all industrial workers in the form of ascribed labor/ 
The machine age posed many new problems both in the agrarian 
and in the industrial spheres of life. The ruling bureaucracy solved 
them — clumsily, no doubt, but successfully insofar as the preserva- 
tion of its hegemony was concerned. The Tsarist regime emanci- 
pated the serfs, but it maintained a tight control over the villages, 
which were administered in a quasi-Oriental manner. During the 
last decades of the 19th century the Russian government, by direct 
and indirect taxes, seems to have taken from the peasants almost 
the whole of their agricultural produce proper — almost 50 per cent 
of the entire peasant income. 28 And the same bureaucracy, which so 
effectively upheld its acquisitive interests, was perfectly willing to let 
the landed aristocracy lose a large part of its estates. Between 1861 
and 1914 the land owned by this group shrank by over 40 per cent. 24 
And Stolypin's reform program of 1906 showed the absolutist 
officialdom considerably more interested in creating a class of strong 
peasant owners than in protecting the landed prerogatives of its 
proprietary wing. 

In the nonagrarian sector of economy the adjustments were simi- 
larly ingenious. The government encouraged private capitalist enter- 
prise in industry and commerce and — to a lesser extent — also in 
communications and banking. But at the beginning of the 20th 
century it managed the bulk of the country's railroads; it maintained 
fiscal control over the comprehensive "monopoly" industries, and it 
occupied a key position in foreign investments. By means of state 
guarantees it influenced something like a third of the nonmonop- 
olized light industry, and in 1914 no less than 90 per cent of the 
core of heavy industry, mining. 25 

These data indicate the strategic position that the Tsarist regime 
occupied in the economy of Russia at the beginning of the 20th 
century. In conformity with the majority of other analysts, the 
prominent Soviet economist, Lyashchenko, notes that the Russian 
banking system prior to the revolution "differed materially from the 

managed industry temporarily shrank (ibid.: 440 ft.), but it rose again impressively 
during the later part of the century. The fourth census reports that for 1781-83 there 
were about a 10,000 "souls" ascribed to the state-owned Mountain Works and 54,000 
"souls" to private units (ibid.: 441). The somewhat less complete report of the 
Manufactures Collegium noted for 1780, 51,000 ascribed "souls" for the private 
Mountain Works and about 24,000 ascribed "souls" outside the key region of Russian 
industry, the "Mountains" (ibid.: 493). 

/. Heavy industry formed the core of the state works, and until "the beginning 
of the nineteenth century, the iron mines and smelting works were manned ex- 
clusively by forced labor" (Mavor, 1925, I: 534). 

CHAPTER 6, C l8l 

banking system of the Western capitalist countries. . . . The state 
bank was the central bank of the entire Russian credit system," and 
the director of the credit department of the treasury "controlled the 
entire financial apparatus of the country." 26 

There is no need to rest the evaluation of Russia's societal order 
on the single criterion of financial control; but it certainly is worth 
noting that one bureau of the Tsarist state apparatus did control 
the country's entire financial system. Considering the role of the 
Tsarist bureaucracy in rural and urban society, it is difficult to 
avoid the decision that even at the beginning of the 20th century 
the men of the state apparatus were stronger than society. 27 

d. Ottoman Turkey 

The later development of Ottoman Turkey combines features of 
the Byzantine and the Russian patterns. The Turkish empire re- 
sembled Byzantium, with whose territory it was largely congruent, 
in that it also originally controlled classical areas of hydraulic 
economy; and it resembled Tsarist Russia in that it was also deeply 
influenced by the industrial society of modern Europe. It differed 
from Byzantium in that the loss of its hydraulic provinces virtually 
coincided with the decline of its political prominence; and it differed 
from Russia in that the growing economic and cultural influence of 
the industrial West was accompanied, and partly preceded, by a 
successful encroachment upon Turkey's sovereignty. 

e. Diversified Final Evolutions 

In all three countries outside aggression was a crucial factor in the 
weakening of the despotic regime; and this indirectly confirms the 
staying power of the Orientally despotic order. 

In the case of Byzantium, it is not entirely clear whether the final 
marasmus of the despotic regime was caused primarily by external 
or internal factors — that is, by the conquest of 1204 or by the ex- 
cessive growth of landlordism. It is clear, however, that the growing 
proprietary forces did not dissociate themselves sharply and creatively 
from the decaying state. The impact of the West was sufficiently 
strong to paralyze the traditional despotic government, but it was 
not strong enough to pave the way for the growth of a new balanced 
and property-based (capitalist) society. 

In the case of Russia, bureaucratic absolutism suffered a mortal 
blow from outside only in 1917. Prior to this date a marginal 
Oriental despotism adjusted itself successfully to the conditions of 


an advancing industrialization. The Tsarist government made more 
and more concessions to mobile and immobile property; and during 
the last period of its existence it even permitted a number of politi- 
cal organizations to operate on a national scale. 28 But these develop- 
ments notwithstanding, the bureaucratic regime perpetuated itself 
until the beginning of the year 1917. 

In the case of Turkey, foreign powers broke the backbone of 
Ottoman independence in a series of wars; and although Russia 
participated in the military defeat of Turkey, Western European 
influence prevailed in the ensuing transformation. It was under 
Western European influence that Turkey undertook important con- 
stitutional reforms. Due to the lesser significance of independent 
proprietary developments both in land and capital, the Turkish 
reforms were at first even more superficial than the reforms accom- 
plished in the Tsarist empire, and this despite the fact that a first 
parliament was established in Turkey as early as 1876/7. But the 
weakness of the independent internal forces was to some degree 
compensated for by the increasing decay of the traditional state 
apparatus, which finally collapsed after the defeats suffered in the 
Second Balkan War and in World War I. 

4. Marginal Agrarian Despotisms Containing 
Conspicuous Hydraulic Elements 

Among marginal agrarian despotisms Muscovite Russia and Middle 
Byzantium, which exhibit numerous cultural similarities, share one 
trait that is particularly relevant to our inquiry: in neither civiliza- 
tion did agrohydraulic operations play a significant role. On the 
other hand, Liao and Maya society, which culturally had little in 
common, are alike in that hydraulic features were clearly apparent 
in both of them. 

a. The Liao Empire 

The Liao empire deserves special attention for a number of reasons. 
It is one of the few Far Eastern societies of conquest in which "bar- 
barian" (pastoral) conquerors — in this case, the Ch'i-tan — ruled over 
part of China without shifting their political center from their Inner 
Asiatic grazing grounds to the subdued (North) Chinese territories. 
Liao is the first of the four great historical Chinese dynasties of 
conquest, the three others being Chin (ruled by the Jurchen), Yuan 
(ruled by the Mongols), and Ch'ing (ruled by the Manchus). Liao in- 
stitutions therefore have significant parallels in the Chin, Yuan, and 

CHAPTER 6, C 183 

Ch'ing dynasties, and it would seem also in other dynasties of con- 
quest and infiltration in China and elsewhere. 47 

During the two hundred years of their rule, the Ch'i-tan acquired 
no real understanding of the potentialities of hydraulic agriculture. 
Instead, and not dissimilar to other mounted "barbarians," they 
eyed with suspicion the irrigated fields which impeded the free 
sweep of their cavalry. 20 The greater part of their agrarian territories, 
however, had a long hydraulic tradition. Canals had been dug and 
rivers diked prior to the establishment of Liao power in North China 
and Manchuria; 30 and the Ch'i-tan conquerors seem to have been 
perfectly willing to preserve this hydraulic heritage. When a flood 
inundated thirty villages in present Hopei, "an imperial decree 
ordered the old canals dredged"; 31 and when in 1074 excessive rains 
threatened the population of the Liao River basin, "the northern 
chancellor [ordered] large-scale mobilization of the able-bodied men 
along the river in order to complete the river dikes." An experienced 
official warned that such "large-scale works" would not be advan- 
tageous at this moment and he asked that the labor corvee be 
stopped. "The imperial court approved it and discontinued the 
work." Subsequent events indicated both the soundness of the 
official's warning — the river caused no calamity — and the dimension 
and weight of the hydraulic corvee: "Along the shores of the river 
for a thousand li there was not a person who was not highly 
pleased." 32 

The Liao government was equally well equipped — and consider- 
ably less reluctant — to employ its manpower for nonhydraulic con- 
structions. Highways were maintained and repaired 33 — once with 
a huge corvee of two hundred thousand men; 3 * chains of fortifica- 
tions were erected along the frontier; 35 and two new capitals and 
many palaces, temples, and tombs were built north of the old seats 
of Chinese culture. 36 Literary descriptions and archaeological finds 
make it clear that the Liao labor service was as effective from the 
standpoint of the rulers as it was onerous from the standpoint of the 
people. 37 

Being great builders, the Liao rulers were also great organizers. 
Their offices registered the population for purposes of taxation, labor 
service, and military recruitment. 38 Their postal system was both 
elaborate and fast. 39 And their army was a well-coordinated fighting 

g. This study was facilitated by the fact that the Chinese subjects of Liao, being 
trained in historiography, recorded the institutions of Liao society more fully than 
the scribes of most other conquest societies of Asia that were dominated by pastoral 
rulers. The reasons for this phenomenon are discussed in Wittfogel, 1949: passim. 


machine. We have reason to believe that Chingis Khan shaped his 
own terrifying military organization after the Liao pattern. 40 

These constructional and organizational developments were sup- 
plemented by genuinely hydraulic methods of acquisition. True, 
some "entrusted" territories delivered only their wine tax to the 
central government; 41 but these regions comprised a mere fraction 
of the realm; 42 and eventually most of them came under full gov- 
ernment control. 43 In the great majority of all administrative sub- 
divisions the state insisted on its subjects paying taxes, 44 just as it 
insisted on their rendering labor and military services. Powerful fam- 
ilies and monasteries sought to have households living on their land 
struck from the public registers, but evidently the state made no con- 
cessions in its claim to tax them. 45 

The final crisis of Liao power has all the earmarks of a dynastic 
crisis under a typical agrarian despotism. Here, as in similar circum- 
stances, the landowners increased their acquisitive 46 but not their or- 
ganizational strength. The collapse of the dynasty led to no property- 
based industrial order. Instead it led to the restoration and rejuvena- 
tion of the old agromanagerial society. 

b. Maya Society 

Maya civilization presents ecological and cultural features that in 
several ways are unique. But these "unique" features overlay con- 
structional, organizational, and acquisitive conditions remarkably 
similar to those of other marginal agromanagerial societies. 

The ancient Maya were spread over a wide area, which comprised 
the greater part of present Guatemala, the western part of the Re- 
public of Honduras, all of British Honduras, and Yucatan. Like most 
of Central America this area has a sharply divided rain year. From 
May to October precipitation is heavy, while during the remaining 
period there is little rain. This dichotomy encouraged elaborate hy- 
draulic developments in territories that border the Lake of Mexico 
and also in several highland regions further to the south, the Maya- 
inhabited zones of Guatemala and Honduras among them. However, 
in large sections of the Maya area geological peculiarities decisively 
shaped and limited hydraulic enterprise. Almost the entire lowland 
plain of Yucatan and a great part of the hill zone between this plain 
and the highlands are composed of an extremely porous mineral: 
limestone; consequently precipitation quickly sinks below an easily 
accessible level. 

A landscape which precludes the formation of rivers and lakes is of 
course entirely unsuitable for irrigation agriculture. Worse. The lack 

CHAPTER 6, C 185 

of natural storage places for drinking water, other than some well- 
like waterholes, presents a serious obstacle for any permanent or pop- 
ulous settlements. Persons desirous of establishing such settlements 
would therefore have to make concerted efforts not for purposes of 
irrigation but for the gathering and preservation of drinking water. 
As a result of such efforts we can expect to find hydraulic installa- 
tions that play only a minor role in other agrarian societies. 

When, in 1519, Cortez briefly visited Yucatan, he noted wells (po- 
zos) and water reservoirs (albercas) in the residential compounds of 
the "nobles." * 7 And in 1566 Landa, in the first systematic description 
of Maya civilization, stressed both the unique water difficulties of 
the area and the way in which moisture was provided "in part by 
industry and in part by nature." 48 It is significant that Landa, like 
the authors of the Relaciones de Yucatan, 11 places the man-made de- 
vices for providing water first. 

The installations for providing drinking water were (1) artificial 
wells (pozos or cenotes in the primary sense of the Maya word), 49 (2) 
cisterns (chultuns), and (3) man-made large reservoirs (aguadas). 
The Relaciones report artificial pozos everywhere in the lowland; 50 
and the early observers fully understood the difficulties of digging 
and maintaining good wells without the aid of metal tools. 51 Even 
after the introduction of iron implements, the maintenance and use 
of the man-made wells often required ingenious communal action. 52 
In some cases the methods employed were intricate "past belief," 53 
involving the active participation of "the population of a city." 54 

But important as the cenotes were, they did not as a rule provide 
water for large populations. Says Casares, a modern Yucatan engi- 
neer: "If we were to depend on the wells only for the supply of 
water, the greater part of our peninsula could not be inhabited." 55 
This being so, the cisterns and aguadas of Yucatan become crucially 

Bottle-shaped subterranean constructions with circular openings, 
chultuns, have been discovered in several places. At Uxmal, Stephens 
noticed "so many of them, and in places where they were so little 
to be expected, that they made rambling out of the cleared paths 
dangerous, and to the last day of our visit we were constantly finding 
new ones." 56 These constructions seem to have provided "immense 
reservoirs for supplying the city with water." 67 

In part. Besides the cenotes and the cisterns,* the ancient Mayas 

h. RY, I: 116, 144, 182, 206, 210, 221, 248, 266. Occasionally major emphasis is 
placed on the natural pozos (ibid.: 47, and perhaps 290). 

1, Stephens (1848, I: 232) assumes that the chultuns of Uxmal had provided water 
for the people of the ruined city — "in part at least." Casares (1907: 227) also com- 


constructed large pools or lakes, aguadas. Even in the hilly regions 
where the terrain provided natural waterholes or cavities, sartenejos, 
Casares considers the aguadas, whether natural or artificial, much 
more important. Those that were man-made differed greatly in 
shape and quality: "Some have a bottom made out of stones and 
some have not such stones, and they are of all sizes — true works of 
art they are — that show the ingenuity and attainments of their 
builders." 5S 

Few students have searched for these aguadas as eagerly as did the 
pioneer explorer, Stephens. At first glance, many of them seemed 
natural, 69 and Stephens' informants felt sure — and recent research 
has proven them to be right 60 — that "hundreds are perhaps now 
buried in the woods, which once furnished this element of life to 
the teeming population of Yucatan." 61 

From the standpoint of hydraulic organization the importance of 
this fact can scarcely be overrated. The cenotes usually required the 
cooperative efforts of smaller communities only; and the urban 
cisterns were probably constructed and maintained by the work 
teams that "built at their own expense the houses of the lords." 62 
But in the case of the aguadas large-scale cooperation was impera- 
tive. In the midnineteenth century a ranchero, who wanted the 
aguada near his estate cleaned, "secured the co-operation of all the 
ranchos and haciendas for leagues around, and at length fairly en- 
listing them all in the task, at one time he had at work fifteen 
hundred Indians, with eighty superintendents." 63 This much co- 
ordinated labor was required when a single aguada had to be 
cleaned with iron tools. Under the stone-age conditions of the 
ancient Maya, the cleaning, and still more the building, of a chain 
of aguadas certainly involved huge work teams. 

Further studies must be made before the institutional weight of 
the man-made cenotes, cisterns, and aguadas can be fully determined. 
But even our present limited knowledge entitles us to state that 
the constructional operations of the Maya include a not inconsider- 
able hydraulic sector. Aguadas were in use not only in the lowlands 
but also in the hill zone, 84 where some of the most ancient centers 
of Maya civilization were located. 85 And irrigation canals, artificial 
lakes, and other familiar types of hydraulic works have been dis- 
covered in the highland sector of the Maya area i and, of course, 
also in the hill zone.* 

ments on the limited capacity of these cisterns to satisfy the water needs of most of 
the ancient cities. 

j. In the old Maya city of Palenque, Stephens discovered the remains of a water 
channel faced with large stones (Stephens, ITCA, II: 321 and 344). Blom found an 
elaborate drainage system "in other parts of the ruins" (Blom and La Farge, TT, I: 

CHAPTER 6, C 187 

The nonhydraulic constructions of the ancient Maya have been 
frequently described. The early Spanish records stress the magnitude 
of the "houses" and "edifices," which the people built for their 
secular and priestly masters; 66 and grandiose ruins confirm the early 
written evidence. Massive stone highways connected a number of 
cities, and like the pyramids, palaces, and temples they must have 
required great levies of corvee labor. 67 

No compensation was given for certain types of the construction 
corvee; 68 and a similar policy may have prevailed also with regard 
to other corvee services, including agricultural labor for "the 
lords." 69 But whether the pay arrangements for labor services were 
uniform or not, there can be little doubt that the commoners worked 
for their masters in a disciplined manner. Prominent men, obviously 
officials, "who were very well obeyed," 70 acted on the ruler's behalf. 
And the power of the sovereign, who controlled either a single city- 
state or a cluster of such units, can be judged from the fact that local 
officials received no share of the tax they collected for delivery to 
the center." 1 The so-called "town councilors," who assisted the highest 
local official, were "in charge of certain subdivisions of the town, 
collecting tribute and attending to other municipal affairs." 71 Ac- 
cording to a regional description, the officials of the town wards 
had "to attend to the tribute and services (communal labor?) at 
the proper time and to assemble the people of their wards for 
banquets and festivals as well as for war." 72 In addition to a variety 
of civil officials, who used a hieroglyphic script and who, among 
other things, kept land records, 73 there were military officials, some 
holding their posts for life, some being appointed for three-year 

189). He also noticed a "fairly elaborate" irrigation system in Amatenango, Chiapas 
(ibid., II: 396), a region which was formerly part of the Old Maya empire. Further 
to the east, in Guatemala, Stephens (ITCA, I: 206) encountered "a large artificial 
lake, made by damming up several streams." A canal in Honduras, probably pre- 
historic, may have "served to irrigate a large portion of the lower plain" near Lake 
Yojoa (Strong, Kidder, and Paul, 1938: 101). 

k. The hill zone, intermediate between the mountain region and Northern Yucatan, 
contains troughlike depressions, whose clay bottoms hold "lakes, swampy lowlands, 
and streams" (Lundell, 1937: 5; Ricketson, 1937: 9; Cooke, 1931: 287), but even here 
the greater part of the terrain is composed of a limestone so porous that the natural 
precipitation quickly sinks below a readily accessible level, creating a dangerous 
deficiency during three or four months of every year (Ricketson, 1937: 10). Bottle- 
shaped chultuns, "excavated in the solid limestone throughout the region," may have 
been used for storing water, if their walls were "rendered impervious by plaster" 
(ibid.: 9 ff.). An aguada near Uaxactun is "doubtless the remains of an ancient reservoir, 
and excavation in its bottom would probably lay bare the stone flooring with which it 
originally had been paved" (Morley, 1938: 139). 

m. The local officials were supported by the people, who worked their fields, main- 
tained their houses, and served them personally (Tozzer, 1941: 62 ft., n. 292; Roys, 
1943: 6a). 


terms. 7 * Picked men, who did most of the fighting and who received 
a special compensation, seem to have constituted cadre troops, but 
"other men could also be called out." 75 The rulers determined (and 
limited) the duration of a campaign in accordance with pragmatic 
considerations, October to the end of January, the agricultural slack 
season, being considered the most suitable time for waging war. 76 

In the acquisitive sphere the power of the regime over its subjects 
was equally unchecked; and there is no reason to doubt that the 
rulers used their opportunities to the full. It has been said that 
"tribute" was light; 77 and the amounts requested from individual 
households may indeed have been modest. But it must be remem- 
bered that under Mexican and Inca dominion, subjects who culti- 
vated the fields for the state and the temples paid no taxes. In 
contrast to this, the Maya commoners who worked the fields of their 
masters delivered in addition "maize, beans, chile, poultry, honey, 
cotton cloth, and game." 78 One regional report implies that such 
tributes were voluntary, but another dealing with the same locality 
notes that anyone who failed to pay would be sacrificed to the gods. 79 

5. "Loose 2" or "Marginal 1"? 

Our survey of Byzantium and Russia and of the Liao empire and 
Maya civilization leads to several conclusions. The hydraulic density 
of the four institutional complexes differs greatly: it is very low 
or zero in the first two cases and relatively high in the last two. 
As a matter of fact, a reasonable argument can be made for classing 
Liao and the Maya as borderline cases of loose hydraulic societies — 
variants of "Loose 2," to use our symbols. For the time being we 
shall view them conservatively as marginal Oriental societies with 
substantial hydraulic elements, "Marginal 1" (M 1), as juxtaposed 
to "Marginal 2" (M 2), that is, Oriental societies with little or no 
hydraulic substance. 

The closeness of M 1 to L 2 and the gap between M 1 and M 2 
are as significant as the fact that all variants of the marginal type 
utilize the organizational and acquisitive methods of despotic state- 
craft. Thus, however marginal they may be hydraulically, their 
methods of social control place all of them definitely in the "Orien- 
tal" world. 

6. Fragmenting Patterns of Inheritance and a 
Government-Dependent Dominant Religion 

Many supplementary data can be adduced to strengthen our basic 
classification. But here we shall refer only to two particularly sig- 

CHAPTER 6, C 189 

nificant criteria: the fragmenting system of inheritance and the 
dependence of religious authority. 

The Justinian Code — Novella 118 — prescribes the equal division 
of property among the children of a deceased person. This provision, 
whatever its origin, fits to perfection the needs of agrarian despotism. 

In Russia proprietary conditions changed as greatly as the in- 
stitutional patterns of which they were a part. Votchina land, a pre- 
Mongol form of strong noble property, was not subjected to frag- 
mentation; and this continued to be the custom until long after 
the noble owners of such land were compelled to serve the state. 
Pomestye land was office land. Originally it passed from father to 
one son; 80 but since all adult males were obliged to render civil or 
military service, the pomestye estate was finally considered a family 
possession to be divided among the father's several heirs. 81 When 
the growing importance of firearms changed the aristocratic cavalry 
army into a plebeian infantry army, fewer noble serving men were 
needed, and Peter I, who merged pomestye and votchina land, made 
the use of the new type of service (state) land hereditary. 82 The law 
of 1731 is an important milestone in the process of making pomestye 
land private. From this year on, pomestye land was divided among 
all the children and, according to the Law Book, "equally among 
all of them." 83 

In Western Europe the nobles emerged from a period of con- 
tractual and limited (feudal) state service with their landed property 
strengthened through primogeniture and entail. Contrary to this, 
and contrary also to the indigenous votchina tradition, the nobles 
of Tsarist Russia emerged from a period of compulsory and un- 
limited state service with their landed property weakened through 
a law of inheritance that prescribed fragmentation. 

In Liao society the ruling tribal stratum — except in the matter of 
imperial succession — seems to have rejected primogeniture, 84 thus 
maintaining its pastoral mores, which permitted all sons to share in 
the family property. In its Chinese sector the regime was careful to 
uphold the traditional Chinese laws. 83 Many edicts praised Chinese 
subjects who conformed to what were considered ideal patterns of 
Chinese familism. 86 This being so, we have no reason to doubt that 
the government also upheld the fragmenting Chinese law of in- 

A fragmenting pattern of inheritance certainly prevailed among 
the Maya. Says Landa: "These Indians did not permit their daughters 
to inherit with their brothers, except it was through kindness or 
good will; and in this case they gave them some part of the accumula- 
tion, and the brothers divided the rest equally, except that to the 


one who had aided the most notably in increasing the property, they 
gave the equivalent." 87 

In Byzantium the Church, being nationally organized from the 
beginning, was well prepared to strive for independence. But the 
rulers of Eastern Rome and Early Byzantium treated religion as 
part of the jus publicum; and even after the catastrophies of the 
7th century, the Byzantine government was able to combat the 
Church's drive for autonomy. In the 10th century the emperor still 
played a decisive role in the selection of the Patriarch. And by 
virtue of his judicial position he could also interfere in church 
administration. 88 

Significantly, the Church became more independent in the last 
phase of the Middle Empire; but even then the emperor could still 
force an obstructing Patriarch to abdicate." It was only after the 
period of the Latin Empire that a completely shattered autocracy 
was compelled to tolerate an almost autonomous Church. 89 

In Tsarist Russia the bureaucratic regime expressed its enormous 
vitality by its victory over the Eastern Church, which after the fall 
of Byzantium shifted its center to Moscow, the "Third Rome." At 
the end of the Mongol period the increasingly powerful Russian 
state exerted an ever-increasing authority over the Church. Ivan III 
seized half the monasterial land in Novgorod; Ivan IV, the Terrible, 
required more taxes and services from Church land; 90 and in 1649 
a new "department of monasteries" further tightened the state's 
control over the Church. 91 In 1721, Peter I abolished the Patriarchate 
and placed the Church under a government body, the Holy Synod. 92 
And a few decades later, in 1 764, the state seized most of the Church 
land without compensation, assigning only one-eighth of the revenue 
from the land to the clergy. 93 In consequence of these combined 
political, religious, and economic measures, "the church became 
more and more a part of the administrative machinery of the state." 94 

In Liao society the problem of an independent Church never 
arose. Government officials, headed by the emperor, shared leader- 
ship in religious ceremonies with a variety of shamans, who, like 
the priests of the Buddhist temples, obviously were not coordinated 
in any nationwide and independent organization ("church"). 95 

The close relation between secular and religious authority among 
the Maya has already been mentioned. The ruler of a territorial state, 
the halach uinic, is believed to have fulfilled "definite religious 
functions"; 9a and certain priests might also be war chiefs. 97 But 

n. A serious conflict was finally decided in favor of the Church, not because the 
Church was such a strong independent factor but because the high bureaucracy 
turned against the sovereign (Ostrogorsky, 1940: 239 ft.). 

CHAPTER 6, C igi 

nothing indicates that the priests of the great temples were bound 
together in any single organization, except insofar as they partici- 
pated in the work of the government. Says Scholes: "In many cases 
priestly and political functions had been combined in such a manner 
that it was difficult, if not impossible, to differentiate them." 98 

7. Location, Genesis, and Institutional Vulnerability 
of Marginal Agrarian Despotisms 

Middle and Late Byzantium, the Liao empire, and the Maya point 
up some of the institutional diversities among marginal agrarian 
despotisms. Discussion of other pertinent civilizations would differ- 
entiate further the picture we have of this significant subtype. The 
Hopi Indians of Arizona, for instance, engage in extremely modest 
hydraulic enterprises — mainly communal spring cleaning" — but 
their building activities are impressive. 

Tibet was faced with certain irrigation tasks in the river valleys 
of the high plateau, 100 but the hydraulic weight of these tasks was 
probably not great. Nevertheless, the "monk officials" 101 did operate 
a well-functioning labor service 102 and an elaborate and fast postal 
system also. 103 Holders of land grants served the government uncon- 
ditionally and as regular officials; 104 and the fiscal apparatus insisted 
on taxing the bulk of the population. 105 

The kings of ancient Asia Minor and certain territorial rulers 
in early China were more outstanding as builders and organizers 
than as hydraulic engineers. But once the common institutional 
denominator is understood, it is easy to recognize that all these civili- 
zations are variants of the marginal type of hydraulic society. 

How did these marginal configurations come into being? And 
how open were they to change? Before trying to answer these ques- 
tions, we must consider their relative location — that is, their spacial 
relation to the major hydraulic areas of the world. 

a. Location 

Taking the major hydraulic zones of the Old and the New World 
as coordinates, we find marginal developments, as for instance the 
nonhydraulic territorial states of ancient China, interspersed between 
definitely hydraulic areas. Many other marginal developments (the 
Hopi Pueblos, the kingdoms of ancient Asia Minor, Middle By- 
zantium, Tibet, Liao, and the Maya) appear at the geographical 
periphery of a hydraulic zone. 

Russia, however, does not. Russia had no close hydraulic neigh- 
bors when, in the 13th century, the Mongols began to introduce 


Orientally despotic methods of government. Cases like Russia are 
more the exception than the rule; but they serve to demonstrate 
that marginal agrarian despotisms may arise at a great distance from 
the nearest conspicuous center of hydraulic life. 

b. Genesis 

The relative location of most marginal agromanagerial states is 
highly suggestive of their origins. The bulk of all such regimes ob- 
viously came into being not earlier — and often demonstrably later 
— than the area's oldest hydraulic civilizations. In some cases, such 
as Byzantium, the marginal territory split off from an older (loose) 
hydraulic complex. In others, the marginal territory was adjacent to 
a hydraulic society proper; and while interrelation cannot always be 
documented, it seems probable that it was the second type which 
stimulated the first. 

The constructional, organizational, and acquisitive patterns of the 
hydraulic center may have been transferred directly to nonhydraulic 
regions during periods of temporary control. Or native leaders may 
have adopted the power techniques of their hydraulic neighbors, 
which from the standpoint of the ruling group had much to recom- 
mend them and which could be easily imposed on a society that 
lacked strong, well-organized, and independent proprietary, military, 
and ideological forces. Or experts in managerial and despotic control 
may have gone from their hydraulic homeland to adjacent non- 
hydraulic territories either in flight or on invitation to become 
teachers or co-leaders in their new environment. 

On an institutional checkerboard familiarity with the hydraulic 
techniques of organization and acquisition was probably all that 
was needed to encourage a changeover from a loosely coordinated 
nonhydraulic tribe to a nonhydraulic managerial community. Thus 
it is easy to understand why the Hopi Indians built fortress-like 
villages similar to those of the more properly hydraulic Pueblos; 
why, like the inhabitants of other Pueblos, they integrated their 
work teams under communal leaders; and why they cultivated the 
fields of their supreme chieftain. 

A combination of state-centered hydraulic and marginal agro- 
managerial societies may emerge from a composite tribal root. In 
prehistoric and protohistoric China such a development may have 
been stimulated by varied and prolonged culture contacts: visits, 
alliances, trade relations, and conquests. 

The introduction of marginal agromanagerial institutions by non- 
agrarian tribal conquerors presents another genetic pattern. In this 

CHAPTER 6, C 193 

case, the conquerors employ and transfer organizational and ac- 
quisitive methods of hydraulic statecraft, although they themselves 
do not, to any relevant extent, practice agriculture, not even in its 
nonhydraulic form. And being nomadic, they may carry these 
methods far beyond the political and cultural borders of any major 
hydraulic area. The Mongol conquest of Russia demonstrates both 

The power of the Ch'i-tan differed from that of the Golden Horde 
in character as well as in origin. The bulk of the agricultural regions 
of the Liao empire had previously been part of the old, loosely 
hydraulic world of China; and the Ch'i-tan masters found it easy to 
perpetuate the traditionally absolutist administration with the aid 
of Chinese officials, who were ready to act as junior partners in a 
somewhat uneasy, but workable, alliance. Like the Mongols of the 
Golden Horde, the Ch'i-tan tribesmen in their great majority re- 
mained pastoralists; but their ruling group integrated itself closely 
with Orientally despotic officials, who directed huge nonhydraulic 
constructions and even considerable hydraulic operations. 

The marginal agromanagerial societies discussed in our survey 
came into being in various ways; but they all seem to have derived 
from compact or loose hydraulic societies. In many instances, such 
an origin is certain, and in others it is likely. But is it the necessary 
and only way? 

By no means. It is entirely possible that some agrodespotic societies 
emerged independently. But obviously we can assume such a develop- 
ment only when the despotic order in question fulfills the organiza- 
tional and acquisitive functions of a hydraulic government and when, 
for geographical and historical reasons, institutional diffusion can be 
excluded as altogether unlikely. Having acknowledged the possibility 
of independent origin, I must add that the cases in which agrodespotic 
regimes in the terms of our inquiry certainly or probably have a 
hydraulic ancestry are so numerous that the cases in which independ- 
ent origins can be established will not substantially change our basic 
contention. Virtually all historically significant agrodespotisms that 

o. The attempt to explain the rise of Muscovite despotism as the consequence of 
external military pressure usually results in the view that this pressure was exerted 
in the main by Eastern nomadic aggressors (see Kluchevsky, HR, II: 319 ff.). The 
imitation of despotic power techniques by a non-"Oriental" government is of 
course conceivable, particularly if the nongovernmental sector of society lacks "strong, 
well-organized, and independent proprietary, military, and ideological forces." How- 
ever, the noble owners of votchina land, although not organized in a corporation, 
were not without strength; and the actual events of the Mongol period show that 
the Great Princes of Moscow, who set out to subdue them, were for a considerable 
time directly under Tatar leadership. 


fulfill no hydraulic functions seem to have been derived from hy- 
draulic societies. 

c. Institutional Vulnerability 

Direct or indirect connection with an agrohydraulic center seems 
to have been necessary for the rise of virtually all marginal agrarian 
despotisms. But a continued connection is not imperative for their 
perpetuation. These marginal agrarian despotisms tend to survive 
even serious internal crises without support from any hydraulic core 
area. However, they are more likely than are the core areas to lose 
their institutional identity under the impact of external nonhy- 
draulic forces. 

Obviously it is enormously difficult to create an effective counter- 
weight to an apparatus government, which has succeeded in repress- 
ing, crippling, and fragmenting those proprietary, military, and 
ideological forces that enabled Medieval (feudal) Europe to evolve 
into an industrial society. Serious political crises occurred in all 
hydraulic societies. But the way in which the men of the apparatus 
overcame them demonstrates the staying power of their methods of 
organization and exploitation. Purposeful political activists strove 
to reestablish the only thoroughly tested type of government, which, 
at the same time, promised them total power and total privilege. And 
their restorative endeavors were greatly facilitated by the political and 
organizational ineptitude of their nongovernmental rivals. Among 
the big landowners, even if they were many, politically ambitious 
elements were much more eager to seize than to restrict total power. 
And the representatives of mobile (capitalist) property, even if they 
were many, were so unaccustomed to think in terms of property- 
based state power that they were satisfied to get on with their busi- 
ness without making the bid for political leadership that was so 
characteristic of the differently conditioned bourgeoisie of the West 

Subjected to the impact of strong external nonhydraulic forces, the 
hydraulic periphery is manifestly more brittle than the hydraulic 
core area. Invaded by nomadic tribes, hydraulic North China at times 
split into several territorial units; but even when the conquering 
"barbarians" became the rulers, these regions maintained their tradi- 
tional agrodespotic power structure. In contrast, the marginal hy- 
draulic society of Western Rome collapsed under tribal attacks, and 
non-Oriental forms of government and society emerged. Also, in 
contemplating the fate of Late Byzantium, it seems legitimate to 
suggest that a more intensely managerial (hydraulic) order would 
have survived the Latin conquest without yielding to the proprietary 
elements within its borders to the point of paralysis. Recent Russia 

CHAPTER 6, D 195 

offers a particularly illuminating example. Shaken but not subdued 
by aggression from the outside, the Tsarist bureaucracy permitted 
the spread of Western ideas, the growth of private enterprises, and 
the establishment of anti-autocratic groups and parties, which, in 
1917, temporarily changed Russia from a single-centered to a multi- 
centered society.* 


1. The Phenomenon 

The effective coordination of absolutist methods of organization and 
acquisition is the minimum requirement for the maintenance of a 
genuine agrarian despotism. Outside this margin we find civiliza- 
tions that, although lacking such a combination, exhibit stray features 
of hydraulic statecraft. The areas in which such stray features occur 
in other societal orders constitute the submarginal zone of the 
hydraulic world. 

2. Cases 

a. Protohistorical Greece 

An institutional analyst of protohistorical Greece cannot fail to be 
struck by the hydraulic quality of Minoan Crete. This civilization 
certainly owed its international prominence to its maritime relations; 
but while acknowledging this, we must not forget that nearness to 
the sea alone explains little. The ancient Cretans, like other seafaring 
peoples, established their thalassocracy on the basis of specific in- 
ternal conditions. 

To what extent Aegean patterns of "fetching water by artificial 
means" and of using canals and ditches for the purposes of refined 
agriculture x made Minoan society hydraulic is not clear. It is clear, 
however, that the islanders accomplished miracles regarding matters 
of drainage and probably also of water supply. 2 We do know that 
Crete was covered with a network of excellent roads. 8 And we have 
reason to believe that the supervisor of public works occupied a 
high position * in the country's complex and centralized adminis- 
tration. 5 The Minoan script is still undeciphered, but the govern- 
ment certainly employed it widely for "bureaucratic methods of 

p. For a fuller discussion of this phenomenon see below, Chap. 10. 


registration and accounting which were handed down from century 
to century and were perfected in the process." G 

These and other facts support the view that "the Minoan civiliza- 
tion was essentially non-European." 7 And although the Minoans 
had too many cultural peculiarities to be called "part of the East," 8 
they were connected through "a few clear and even close bonds with 
Asia Minor, Syria and Egypt." 9 Ehrenberg concludes that "in partic- 
ular the sultan-like life of the kings of Cnossus and Phaestus, their 
courts, their officials, their economy, displayed features which were 
similar to those of their opposite numbers in the Near East; they 
were equally unlike anything Western." 10 

The proto-Greek Mycenaean civilization, which rose when Minoan 
power decayed, accounts for significant quasihydraulic developments 
in Argolis and Boeotia, and probably also in other parts of eastern 
Greece. Between the middle and the close of the second millennium 
b.c. Mycenaean engineers executed great drainage works around the 
Lake of Copais in Boeotia; and they covered Argolis with an elab- 
orate network of roads. 11 Their rulers lived in huge castle-like 
edifices, and they erected monumental tombs. 12 Bengtson compares 
their constructional achievements to "the great creations of the 
ancient Orient, the pyramids and the ziqqurats." 13 True, we hear 
nothing of a bureaucracy, and the use of the early script seems to 
have been restricted. 14 But despite such limitations, Bengtson be- 
lieves that "only a strong central power could plan and execute these 
works," which, considering their magnitude, in all probability re- 
quired the services of both native corvee laborers and captured 

Moreover, an Oriental origin has been suggested for the worship 
of the earth gods and the stars which the historical Greeks inherited 
from their Mycenaean ancestors, and it was indeed in connection 
with such religious observances that they practiced prostration. 15 
But when the Greeks of the classical period refused to perform before 
an Oriental despot the act of submission they considered appropriate 
to the gods, 16 they demonstrated that even if Mycenaean Greece was 
marginally hydraulic, post-Mycenaean Greece belonged to the sub- 
marginal zone of the hydraulic world. In the classical period also 
the monumental edifices of Argolis 1T had long lost their significance; 
and the grandiose temple city of Athens, the Acropolis, whose be- 
ginnings go back to Mycenaean times, 18 was administered by a gov- 
ernment that delegated even the management of its public works to 
private entrepreneurs. 19 

a. Bengtson, 1950: 41. Bengtson mentions the slaves before he mentions the native 
corvee laborers, but he calls the latter as numerous as the former. 

CHAPTER 6, D 197 

b. Early Rome 

Prior to Roman times the Etruscans, who apparently came from 
the marginal hydraulic zone of Asia Minor, 20 are known to have 
engaged in stupendous building activities. Their waterworks in the 
Po Plain are impressive, 21 and others undertaken in Central Italy 
are equally worthy of attention. 22 While under Etruscan dominion, 
the Romans learned how to construct "monumental works." 23 Later, 
but before they established their first colony on Hellenistic ground, 
they began to build solid overland roads. 24 But although such de- 
velopments are more characteristic of a hydraulic than a relatively 
simple rainfall-based agrarian order, Rome at this period manifestly 
was an aristocratic variant of a multicentered non-Oriental society. 

c. Japan 

In ancient Greece and Rome Oriental elements have often been 
overlooked. In Japan they have frequently been overestimated, and 
this for a good reason. Japan is part of the Asian continent, and 
Japanese civilization shares important features with China and India. 
Furthermore, the Japanese have developed one of the most subtle 
systems of irrigation farming known to man. Nevertheless, Japanese 
society never was hydraulic in the terms of our inquiry. 

Why did Japan's rice economy not depend on large and govern- 
ment-directed water works? Any competent economic geographer 
can answer this question. The peculiarities of the country's water 
supply neither necessitated nor favored substantial government- 
directed works. Innumerable mountain ranges compartmentalized 
the great Far Eastern islands; and their broken relief encouraged a 
fragmented (hydroagricultural) rather than a coordinated (hydraulic) 
pattern of irrigation farming and flood control. According to the in- 
stitutional historian, Asakawa, the Japanese landscape permitted "no 
extensive Bewasserungskultur as in Egypt and in parts of western 
Asia and China." 2S Japan's irrigation agriculture was managed by 
local rather than by regional or national leaders; and hydraulic 
trends were conspicuous only on a local scale and during the first 
phase of the country's documented history. 

The rulers of the dominant political center effected a loose politi- 
cal unification at a rather early date, but they were not faced with 
hydraulic tasks that required the coordinated operation of large 
corvee teams. Nor were they conquered by the forces of an Orientally 
despotic state. They therefore failed to establish a comprehensive 
managerial and acquisitive bureaucracy capable of controlling the 


nongovernmental forces of society as did the men of the apparatus 
on the Chinese mainland. 

The attempt to establish a centralized and bureaucratic despotism 
in Japan reached its first spectacular climax in the Taikwa Reform 
of 646. From the standpoint of our key criteria, its objectives can be 
listed as follows: 

I. Construction 

A. Hydraulic. An edict of 646 demanded uniform proce- 
dures relating to dikes and canals. 26 

B. Nonhydraulic. The basic reform edict ordered the 
creation of a system of roads for the imperial post. 

II. Organization 

A. The population was to be counted periodically and 
census registers were to be kept. 

B. A government corvee replaced older local (and quasi- 
feudal) obligations. 

C. A state post was to be operated. 
III. Acquisition 

A. The peasants were to be taxed on the basis of the land 
which the government assigned to them. 

B. Service in the state corvee could be commuted by the 
payment of a tax. 27 

C. A number of officials, particularly local and high- 
ranking dignitaries, were to be supported from land 
holdings, which had often been previously owned by the 
new appointees and which were tax exempt. 

Compared with the Merovingian and Carolingian attempts at ab- 
solute rule, the Japanese program of 646 was much more Oriental. 
This fact cannot be explained by Japanese contact with T'ang China 
alone. For centuries the Japanese had practiced irrigation farming 28 
and their rulers had engaged in constructing works of a non- 
hydraulic type. Thus the effort of the masters of the Reform govern- 
ment to do as the Chinese emperors did was rooted in indigenous 
trends that were definitely, if rudimentarily, hydraulic. 

But these quasi-Oriental trends were unable to shape Japanese 
society. The hydraulic innovations suggested in the Reform lacked 
the dynamism that characterized similar attempts in early hydraulic 
societies. The Reform favored the execution of "public works"; 
but while three of the six T'ang ministries (taxation, war, and 
justice) were taken over with little modification and two others 
(administrative personnel and rites) were successfully modified, the 

CHAPTER 6, D 199 

sixth (the Board of Public Works) found no counterpart in the 
new Japanese set-up. 29 

This omission was no accident. A canal that was dug in 656 struck 
the people as "mad"; and its critics compared it with a useless colossal 
hill that was built at the same time. 30 Moreover, the decrees that 
proclaimed a universal state labor service required many less days 
of corvee work than did T'ang regulations. And the provisions for 
commuting the corvee by paying a tax showed the Japanese govern- 
ment more interested in revenue than in labor. 31 

The assignment (and/or reassignment) of tax-free land to im- 
portant officials was perhaps the Reform government's greatest con- 
cession to the feudal forces of Japanese society. Behind the new 
bureaucratic facade a fierce fight was being made to extend and 
consolidate tax-free land. And so successful were the representatives 
of the centrifugal forces that the official grantees eventually estab- 
lished themselves as hereditary landowners who, like their European 
counterparts, introduced a single-heir system of succession. 32 

As the system of tenure changed, universal census taking collapsed; 
and attempts to reestablish it led nowhere. 33 General taxation met 
the same fate. Many elements of Chinese culture notwithstanding, 
the decentralized and property-based society of the Japanese Middle 
Ages resembled much more closely the feudal order of the remote 
European world than the hydraulic patterns of nearby China. The 
poets of feudal Japan, like their confreres in feudal Europe, glorified 
the heroic deeds of individual warriors or groups of warriors. But 
the loosely agglomerated armies of Medieval Japan did not stimulate 
tactical or strategic thinking. The Japanese writers of the period 
quoted Chinese military authorities, such as Sun Tzu; but feudal 
Japan, like feudal Europe, failed to develop the art of war. 6 Prior to 
1543, the Japanese armies "were made up of small, independent 
bands of soldiers who fought more as individuals than as units of a 
tactical formation." " 

b. The reader will remember that the term "art of war" connotes the practice and 
theory of strategy and tactics. A recent survey of ancient and medieval military or- 
ganization ascribes "the beginnings of an acknowledged art of war" in postfeudal 
Europe to Maurice of Nassau (Atkinson, 1910: 599), who played a decisive role in the 
latter part of the Dutch War of Independence. 

c. Brown, 1948: 236 if. A collection of early Japanese texts, Gunsho Ruiju, contains 
many references to Sun Tzu and other military theoreticians of his period. But the 
Japanese treatment of warfare is "a rather scattered melange quite unlike Sun Tzu. 
. . . The first integrated treatment of the subject comes in a work by Takeda Shingen 
(1521-1573)" (from a letter of February 16, 1954, from Dr. Marius Jansen, University of 
Washington, Seattle, who established this point in collaboration with his colleague, 
Dr. Richard N. McKinnon) 


The absolutist concentration of government power, which char- 
acterized the Tokugawa period (1603-1867), again resembled more 
closely Western absolutist developments, both in its economic aspect 
(the slow rise of property-based commercial and industrial capi- 
talism) and in its political limitations. It was during this period — 
actually in 1726 — that "the first tolerably exact census" was taken. 34 
It was then that the road system spread vigorously; 35 and it was 
then that the government, like certain of the prominent feudal lords, 
dug a number of locally important canals. 36 

But despite these and other activities — which, except for the irri- 
gation works, find illuminating parallels in absolutist Europe — the 
absolutist regime of Japan was not strong enough to establish its 
acquisitive power over the whole empire. Out of a national revenue 
of twenty-eight or twenty-nine million koku, the representatives of 
supreme power, the Tokugawa shoguns and the court, arrogated to 
themselves only about eight million koku, while by far the larger 
part of the revenue remained in the hands of great feudal vassals. 37 
Japanese absolutism sharply restricted the power of the feudal lords. 
But until 1867 it was unable to eliminate them. 

While stressing the similarities between traditional Japanese 
society and the feudal and postfeudal West, we must be careful not 
to oversimplify the picture. The Oriental quality of many Japanese 
institutions and ideas is beyond doubt. On the lower and local level, 
Japanese irrigation agriculture required quasihydraulic coordina- 
tion and subordination; and the feudal lords' insistence upon ab- 
solute obedience may, at least in part, reflect such quasihydraulic 
relations. Rudiments of a postal system seem to have existed prior 
to the Tokugawa period; 38 and the symbol of total submission, 
prostration, persisted until modern times. 4 * The members of the 
ruling group, although strongly imbued with a military spirit, con- 
tinued to think in terms of a somewhat adjusted Confucianism; S9 
and although they invented simplified phonetic symbols, they em- 
ployed with genuine pride the Chinese script, which, like Con- 
fucius' conception of the gentleman-bureaucrat, was better suited to 
a civil and learned officialdom than to a war-minded knighthood. 

To sum up: traditional Japan was more than Western feudalism 
with wet feet. While the Far Eastern island society gave birth to a 
property-based and genuinely feudal order, its many and cherished 
elements of Chinese policy and thought show that, in a submarginal 
way, it was related to the institutional patterns of the hydraulic 

d. During my stay in Japan in 1935 a number of university professors greeted each 
other in my presence — and prior to an official banquet — by prostration. 

CHAPTER 6, D 201 

d. Pre-Mongol (Kievan) Russia 

Russian society prior to the Mongol conquest (1237-40) presents 
another and equally illuminating aspect of the hydraulic submargin. 
In pre-Kievan and Kievan days the subsistence economy of the "Rus" 
included stock-raising; e but its mainstay was agriculture, rainfall 
agriculture. 40 Under the conditions of a primarily natural economy 
this agriculture favored the development of a broadly spread landed 
nobility, which was subordinated to the territorial princes in a loose 
way/ Below this stratum, but above the slave-like kholopi* 1 a class 
of free cultivators moved with comparative ease; 42 and the towns- 
people were even less restricted. Their "council," the veche, could 
take independent political action not only in the great northern 
republic of Novgorod 43 but also in such capital cities as Vladimer, 44 
and even in Kiev. 45 Prior to the establishment of the Kievan state 
(ca. 880) ° legal transactions could be consummated, and without in- 
terference from any princely authority, by the heads of the rural — 
and urban — communities, which in the most ancient Russian law 
code extant are called mir. n And even in the Kievan era (ca. 880- 
1169), the government, although considerably stronger than pre- 
viously, was far from being absolutist — indeed as far from such a 
condition as the government of any feudal state in the contemporary 
West. Institutionally speaking, Kievan society manifestly belonged 
to the protofeudal and feudal world of Europe. 

It belonged to this world, but in a way that requires special 
investigation. Like hydraulic society, feudal society, too, has an 
institutional margin; and Russia's tribal civilization, which arose 
on the eastern periphery of the feudal world, was for centuries, and 
particularly after 88o, 4C dominated by the Varangians, 47 who were 
rooted in — and repeatedly supported by — a northern fringe area: 
Scandinavia. But although Rurik had once received a fief from 
the Frankish emperor, 48 he did not impose the Western European 
system of land tenure on the Eastern Slavs. Nor did his successors. 

e. The oldest known version of the Russian law, Russkaya Pravda, mentions crimes 
pertaining to oxen, sheep, goats, horses, calves, and lambs (Goetz, RR, I: 15 ff.). 

/. This fact has been established through the pioneer investigations of Pavlo- 
Silvansky. For a survey of his major conclusions see Borosdin, 1908: 577. For an inde- 
pendent study arriving at similar conclusions for the early Russian society see Hotzsch, 
1912: 544. 

g. Vernadsky (1943: 368) places Oleg's conquest of Kiev "between a.d. 878 and 880 
(tentatively, 878)." 

h. Russkaya Pravda, I, 17 = Goetz, RR, I: 8, 9. Cf. Vernadsky, 1948: 134. In the 
third version of the Law, the early term, tnir, is replaced by gorod, city (Russkaya 
Pravda, III: 40 = Goetz, RR, I: 28, 29, cf. 272 ff.). 


The native nobles and the members of the princely retinue, the 
druzhina, operated under no feudal contract. 49 Their freedom to 
"ride away" 60 indicates a type of independence that in Western 
feudalism was more the exception than the rule. 61 On the other 
hand, the princely rulers of the various territorial states drew their 
maintenance not from royal domains, as was usual in most feudal 
countries, but from a general tax, custom fees, and legal fines. 62 

Thus Kievan society resembled the feudal order of the West in 
that the rulers shared the power of making political decisions "with 
the popular assembly (veche) and the senate (boyarskaya duma)"; 63 
and the nobles were able to establish a form of absolute landowner- 
ship that the lords of Western Europe matched only at the close of 
the Middle Ages. As in the feudal West, the cities — at least the large 
ones — and the nobles paid no taxes. 64 But this extremely loose ar- 
rangement interlocked with a fiscal system that permitted the sover- 
eign to tax the entire rural population. The principle of levying a 
tax on each fireplace was employed in Byzantium; 65 and the semi- 
pastoral Khazars applied it to those Eastern Slavs over whom, prior 
to the victory of the Varangians, they had control. The Varangians 
followed the fiscal procedure of the Khazars, 66 and they continued to 
do so with modifications during the whole Kievan period. 67 They 
also adopted other "Asiatic" features from the Khazars or related 
tribes. For a time their rulers referred to themselves as "khagans"; * 
and prior to the introduction of Christianity they apparently kept 
their numerous concubines in harem-like confinement.* 

Direct Byzantine influence made itself felt relatively early. In 
addition to many literary and artistic elements, the Russians adopted 
Eastern Christianity and Byzantine law, both of which affected the 
political climate of Kiev. The Byzantine ("Greek") priests, who 
came to Russia, carried with them significant ideas of theocratic 
rule and subordination. Accustomed to act as part rather than as 
rivals of the secular government, they certainly enhanced the power 
of the prince.* The introduction of Byzantine law further strength- 

i. Vernadsky (1943: 282) assumes borrowing from the Khazars. The title khaghan was 
borne by "the first Kievan princes." Apart from Vladimir, his son Yaroslav is also 
known to have been thus addressed by the Metropolitan Hilarion (ibid.: 370, and n. 

;'. Prior to his conversion, Vladimir is said to have had about 800 concubines 
(Nestor, 1931: 55). 

k. This fact has been stressed by a number of historians. Platonov points out that 
the "Christian and Byzantine conception of the prince as a ruler by divine right . . . 
was opposed to the pagan view that the prince was a mere leader of a druzhina, and 
could be driven out and killed" (Platonov, 1925: 40). The Soviet academician Grekov 
quotes fully the pertinent statement in the Nestor Chronicle: "God gives the power to 

CHAPTER 6, D 203 

ened the Kievan sovereigns. In the second Constantinople-influenced 
version of the Russian law the ruler and his functionaries emerge 
clearly as the possessors of supreme judicial authority. 68 

But Kievan society did not accept the legal notions of the great 
Eastern empire in toto. The Byzantine code prescribed corporal 
punishment for horse stealing; but the revised Russian law continued 
to demand a fine for this act. 59 Despite its great prestige, Byzantine 
law did not supersede the Kievan view that a free man should not 
be beaten. 

3. Comment 

Evidently, the civilizations in the submargin of hydraulic society 
exhibit a wide institutional range; and their basic structures can 
be understood only if they are viewed first in their primary institu- 
tional context. However, certain secondary qualities, which link them 
to the hydraulic world, must not be overlooked: 

1) A civilization that was once part of this world may, in a later 
nonhydraulic phase, still preserve certain traces of its previous con- 
dition, which, although not necessary to the new configuration, are 
compatible with it. Post-Mycenaean Greece probably belongs to this 

2) The voluntary adoption of desirable "Oriental" features ac- 
counts for such phenomena as Taikwa Japan and Kievan Russia. 

Another point that is valid for marginal hydraulic societies is 
also valid for the submargin. It would be incorrect to view as sub- 
marginally hydraulic an agrarian society which exhibits certain 
despotic features of organization and acquisition but which has no 
known link to the hydraulic world. Individual features of hydraulic 
statecraft, such as the levying of a general tax or the collection of a 
general tribute, certainly have emerged in civilizations which had 
little or no contact with this world. In a number of tribal societies 
this obviously happened; and if we did not know of the Asiatic 
background of the Khazars, we might also feel tempted to place 
their system of tribute gathering in this independent and residual 
category. Comparative analysis must in each instance decide whether 
we are dealing with submarginally hydraulic or independent trends. 

whomever he wishes; the Supreme Being appoints whomever he desires as the caesar 
or prince." Each state should be headed by a caesar or prince, and state power is of 
divine origin — these are indeed "the familiar features of the Byzantine conception 
of state power." Grekov underlines the authoritarian spirit of the famous Christian 
chronicle: "Anyone who attacked the authority — according to the theory — opposed 
God." And "Yaroslav's merit lies in the restoration of a single authority in the state" 
(Grekov, 1947: 133 ft.)- 



The submarginal zone of the hydraulic world cannot be explained 
by a simple formula. Nor is it necessarily self-perpetuating. A num- 
ber of historically prominent civilizations of the submargin have 
crossed the institutional divide and become either marginal hydrau- 
lic societies or hydraulic societies proper. Others have moved in the 
opposite direction. 

The civilizations discussed so far have been essentially agrarian. 
The very concept of a hydraulic economy implies agriculture. But 
the history of the Ch'i-tan, the Mongols, and other tribal conquerors 
demonstrates that Oriental despotism is not confined to agrarian 
societies. Nonagricultural peoples, too, may adopt and transmit 
techniques of despotic government; and they may "Orientalize" non- 
agricultural as well as agricultural groups. The importance of this 
fact for the understanding of many despotic conquest societies and 
of the dynamics of the institutional divide is obvious. 

1. Nonagricultural Peoples Adopting and Transmit- 
ting Power Devices of Agrarian Despotism 

Representatives of many extractive modes of subsistence — 
gathering, hunting, and fishing — have lived at the fringe of the hy- 
draulic world. In this respect, the margin of Pueblo society x and 
the early phases of Aztec history are instructive. But no primitive 
nonagricultural group has played as important a role as the pas- 
toralists. The New World lacked animals suitable for drawing carts 
and carrying men. The Old World had several species that could be 
so used. Their domestication greatly benefited the plant breeders; 
but primarily it benefited the pastoralists, who, after the invention of 
riding, became the military equals, and at times the masters, of large 
and wealthy agrarian commonwealths. 2 

a. Such Devices Not Necessary for, but Compatible with, 
Nomadic Pastoralism 

Pastoral nomads frequently supplement their herding economy 
by farming. 3 Yet the need to move their herds prevents them from 
giving more than casual attention to whatever crops they plant near 
their camping grounds. Their migratory way of life, however well 
regulated, excludes the construction of elaborate and permanent 
works of water control, which form the foundation of hydraulic 

CHAPTER 6, E 205 

But this mode of life does not prevent them from adopting 
Orientally despotic methods of organization and acquisition. To be 
sure, such methods do not grow out of the needs of pastoral life. 
Although some coordination and subordination are imperative for 
effective camping and trekking, and although disciplined procedure 
is highly advantageous for hunting and warfare,* these practices do 
not necessarily lead to the establishment of a political apparatus 
stronger than all nongovernmental forces of society. Technical 
factors (the ever-recurring need for dispersing herds and men) and 
social factors (the resistance of the free tribesmen to the demand 
for total submission) work in the opposite direction. Even subordina- 
tion under a strong military leader is essentially voluntary. Limited 
in time and not bulwarked by irreversible organizational arrange- 
ments, it rarely, if ever, destroys the loose and fluid character of the 
tribal society. 6 

The chiefly leader and those close to him are eager to place 
themselves in a position of permanent and total power; but as a rule 
they attain this goal only after submission to, or conquest of, a 
hydraulic country. In the first case the overlords of the agrarian 
state may apply their own patterns of political control (registration, 
corvee, taxation) to the submitting herders, whose chieftain usually 
emerges as the absolute and permanent master of his tribe. In the 
second case the supreme chieftain (khan, khaghan, etc.) seizes the 
power devices of the agromanagerial civilizations he has conquered. 
Bulwarked by indigenous officials who maintain the traditional ad- 
ministration and by a group of tribal followers whose number grows 
with his successes, he reduces his noble rivals to a shadow of their 
former importance, if he does not annihilate them altogether. 

In both cases the tribesmen may lose their cultural — and eventually 
also their sociopolitical — identity. This happened to many Arab 
groups under the Abbassid caliphate. In such a situation the prob- 
lem itself ceases to exist. However, submitting tribesmen are usually 
not eager to relinquish their old way of life; nor are tribal con- 
querors as easily absorbed as legend has it. 6 With proper modifica- 
tions, the tribal masters of a compound hydraulic empire may main- 
tain their social and cultural identity; and while doing so, they may 
impose their newly acquired power techniques to outlying nonhy- 
draulic countries. This happened when the Mongols, after the con- 
quest of North China, subdued Russia. 

The disintegration of a compound hydraulic empire may again 
make all or some of its tribal elements autonomous; and it is at this 
moment that the perpetuation of despotic power, under conditions 
of tribal pastoralism, is put to the test. At times the despotic regime 


dissolves as completely as the empire of which it was a part. But 
historical experience shows that the beneficiaries of absolutist gov- 
ernment continued in a privileged position, at least to some extent 
and for some time. Obviously then, despotic methods of organization 
and acquisition, although not a necessary adjunct of nomadic pas- 
toralism, are definitely compatible with it. 

b. The Brittleness of Orientally Despotic Power at the 
Pastoral Fringe of the Hydraulic World 

Recent studies have provided a wealth of data concerning all these 
processes for the Ch'i-tan tribes, who, as Liao rulers, were the 
temporary masters of the northeastern fringe of China. Many mono- 
graphs have clarified corresponding aspects of Mongol history; and 
future investigations of the tribal conquest societies of the Near 
East, of Persia, India, and pre-Spanish America will certainly bring 
to light many other varieties of this important institutional con- 

Already our present knowledge enables us to juxtapose the pastoral 
and the agrarian forms of a marginal hydraulic society. Without 
doubt, the staying power of genuine despotism is much greater 
under agricultural than under tribal, pastoral, or nomadic condi- 
tions. The fluidity of a steppe economy encourages diffusion and 
separation and, as a corollary, the growth of independent centers of 
animal wealth and military power. Natural calamities or serious 
military reverses weaken and dissolve a pastoral despotism as quickly 
as the fortunes of war and conquest bring it into being. The 
meteoric rise and fall of many steppe empires in Inner and West 
Asia and in Southeast Europe illuminate the brittleness of pastoral 

The "Black" Ch'i-tan tribes, who grazed their herds in Northern 
Mongolia a hundred years after Liao fell, revealed few traces of the 
coordinated political order maintained by their forebears either in 
the Far East or in Turkestan. 7 After the collapse of the Great Khan's 
empire, Mongol power shrank to a shadow of its former self, but it 
did not disappear altogether. In 1640 the Mongol-Oirat were still 
restrained by laws which, although considerably milder than Chingis 
Khan's Yasa* forced the tribesmen to participate in a relatively 
heavy transport corvee. 9 Manifestly, postempire Mongol society was 
not entirely lacking in cohesiveness when attachment to the rising 
Manchu star gave their secular and religious masters a chance to sup- 
port, in a privileged if secondary way, another ambitious attempt 

CHAPTER 6, E 207 

to establish a despotic regime, first in the margin and later in a great 
core area of the hydraulic world. 

2. Agricultural Civilizations Crossing the 
Institutional Divide 

The changeover of pastoral societies from a nonhydraulic to a 
hydraulic order proceeds, as a rule, on a geographical as well as on an 
institutional level. In contrast to this, changing agrarian societies do 
not change their locale. They move from one order to another ex- 
clusively on the institutional level. 

A second difference concerns the potential range of the change- 
over. Pastoral societies, which preserve their economic identity, may 
shift from the submarginal to the marginal zone of the hydraulic 
world and vice versa. Agrarian societies that were originally sub- 
marginal may become marginal hydraulic or full-fledged hydraulic 
societies and vice versa. 

Like pastoral societies, agrarian societies change their institutional 
quality most frequently at the geographical periphery of agroman- 
agerial areas; for it is here that the forces of the hydraulic and the 
nonhydraulic world have wrestled with each other for millennia. The 
societal transmutations of Greece, Rome, Spain, and Russia are all 
part of this gigantic interaction. 

a. Greece 

From a marginal or submarginal hydraulic position, Mycenaean 
Greece evolved into a civilization whose aristocratic and democratic 
energies prevented the state from exerting unchecked control over 
the nongovernmental forces of society. The Greeks of Homer, Hesiod, 
and Sophocles prostrated before certain of their gods; but they re- 
fused to recognize the supreme representative of state power as their 
master (despotes). 

For many centuries, and despite their proximity to the hydraulic 
world, the Greek cities in Western Asia upheld within their limits 
the principles of a multicentered society. Only in the wake of Alex- 
ander's conquests did the old constitutional freedoms begin to shrink. 
The Hellenistic sovereigns of the Orient reduced the political inde- 
pendence of their own co-nationals in Asia and at home. Together 
with their Macedonian-Greek aides, they readily donned the robes 
of Orientally despotic power. 

The early Roman empire and Byzantium completed what the 
Hellenistic dynasties had initiated. The Greeks of the Near East — 


and those of the motherland — became part of a hydraulic empire, 
which included impressive areas of loose (Syria) and compact (Egypt) 
hydraulic economy. During the 7th century this empire shifted to 
the margin of the hydraulic world. Later the conquering Turks re- 
stored it once more to a loosely hydraulic position. 

The Byzantine and Turkish Greeks were no longer the Hellenes 
of Hesiod, Pericles, and Aristotle. This is probably true ethnically, 
and it is certainly true institutionally. The scions of Mycenae, who 
during the classical period and for the free members of their com- 
munity created exemplary models of democratic citizenship, were 
the ancestors of the Byzantine Greeks, whose elaborate court cere- 
monial made "Byzantinism" a catchword for man's total, if ritualized, 
submission to total power. 

b. Rome 


I n Greece the shift to hydraulic forms of state and society was initi- 
ated by Alexander's conquest. In Rome the establishment of absolute 
and monarchic rule by Augustus signals not the beginning but a rela- 
tively advanced stage of a process that had been under way for about 
two hundred years. 

In the institutional history of Rome the year 2 1 1 B.C. is a fateful 
date. It was in this year that in the subdued Sicilian kingdom of 
Syracuse the Romans "encountered for the first time a subtly elabo- 
rate legal system of a primary agrarian state patterned after Egyptian 
and general Hellenistic models." 10 The victorious Italian republic 
made this system, the so-called Lex Hieronica, "the basis for the 
organization of its first provincial economy." xl By so doing, it 
adopted a basic principle of Hellenistic statecraft, which declared the 
state the holder of absolute power and the owner of all land. 12 

As the successors of Hieron, the Roman conquerors made their 
state, the populus Romanus, the supreme master of Sicily's agrarian 
economy. And they acted similarly also in the other territories of 
their growing empire. In the regions of the Eastern Mediterranean 
this involved little change. But in the western areas of Roman ex- 
pansion nonhydraulic conditions prevailed. It is therefore extremely 
significant that the Italian conquerors, with proper modifications, 
transferred the Hellenistic system "also to the West." 13 

From the Roman point of view the Hellenistic principle of general 
taxation was "a complete innovation." And this innovation was a 

CHAPTER 6, E 20g 

success because it was supplemented by a periodic and comprehensive 
census. According to Hieron's plan, which the Romans adopted, "it 
was the duty of the city magistrates every year to take a census of all 
the farmers of the district . . . recording both the complete acreage 
. . . and the acreage of each crop actually under cultivation." 14 

These external developments did not automatically create a state 
stronger than society in the Roman homeland; but the metropolis 
underwent internal changes which devastatingly weakened the tradi- 
tional aristocratic republic. On the one hand, the unending wars of 
conquest enriched the senatorial landlords, who employed an ever- 
increasing number of slaves; on the other hand, these wars exhausted 
the peasantry. Together with the land-hungry veterans, the im- 
poverished peasants offered an ideal mass basis for the policies of the 
populates and of the victorious generals, who did not hesitate to 
confiscate and redistribute the estates of their erstwhile opponents. 15 
The civil wars also increased the vulnerability of the wealthy busi- 
nessmen, the equites, some of whom as tax farmers, publicani, prof- 
ited greatly from the growth of the Roman realm. During the ad- 
vancing crisis the equites enjoyed as little personal and proprietary 
safety as did the members of the senatorial group. 

Evidently, the internal changes were so closely tied up with the 
country's territorial expansion that any attempt to explain the fall 
of the republic exclusively on the basis of either internal or external 
factors must prove inadequate. The generals who dominated the 
political scene, particularly in the 1st century b.c, rose to power 
because of the size and peculiarity of the territories they occupied. 
It was in these areas that they secured their material support; and 
it was in these areas that they tested the effectiveness of Hellenistic 
methods of government. 

How much did any single individual contribute to the changes 
that occurred in Roman society? For the purpose of our inquiry it 
is sufficient to note that in Caesar's time the senate had already lost 
both its social homogeneity and its uncontested political hegemony 
and that Caesar, who like other great politician-generals of the period 
gave land to the veterans, challenged the senatorial representatives 
of large landed property as a "man of the people," a popularis. Here, 
as elsewhere, absolute power was established through the agency 
of men who used a popular cause to advance their political aims. 

At the time of Caesar's assassination the strongest proprietary force 
in Rome, the senatorial group, had been so shaken that Augustus, 
who officially controlled a number of "imperial" provinces (among 
them the old hydraulic areas of Egypt and Syria) was able to control 
the "senatorial" provinces too. 16 From 29 b.c. on, the senators, who 


previously had been the decisive force behind the administration, 
had to get a permit from Augustus before they could leave Italy; 
and "if the object of their travel was a visit to Egypt, [the request] 
was refused on principle." 17 During the subsequent period the once 
dominant aristocratic landowning senators were more and more 
replaced by persons who became members of the senate because 
they were in the emperor's service. And the representatives of mobile 
wealth and capitalist enterprise, who, as publicani, had collected 
taxes and customs fees for the government and, as contractors, had 
executed certain "public works," were plundered by Pompey, weak- 
ened by Caesar, and subordinated by Augustus. 18 Eventually they 
lost their significance altogether. 19 Thus the Roman metropolis, 
which temporarily had ruled a huge Hellenistically hydraulic em- 
pire without itself being hydraulic, eventually caved in under the 
hammer blows of forces which drew their ultimate strength from this 
very empire. 

In this gigantic process of transformation Augustus was not only 
the grave digger of the old social forces, but also the pioneer in ad- 
ministrative and managerial change. Despite great loyalty to the 
cultural values of Rome, the first emperor (princeps) patterned his 
absolutist state not on early Rome or classical Greece — from which, 
indeed, he would have gotten little inspiration — but on the Hellen- 
istic Orient. 6 By laying the foundations for a salaried officialdom, 20 
he initiated a bureaucratic development that rapidly gained mo- 
mentum in the ist century a.d. 21 

Agromanagerial methods of acquisition and organization had al- 
ready been employed in the provinces under the Republic; now 
they were elaborated and systematized. Confiscations became a stand- 
ard feature of the empire's economic and political life. General 
taxation was bulwarked by the periodic registration of the popula- 
tion, which under Augustus became regular administrative proce- 
dure. 22 Initiating the great nonhydraulic constructions that are still 
associated with the name of Rome, Augustus started to build a truly 
agromanagerial system of roads. He established the state post, the 

a. Of course, the Roman metropolis was not hermetically sealed off from its Oriental 
environment. The growing influence of Hellenistic statecraft was significantly accom- 
panied by the growing influence of Eastern religion, art, technology, and customs. The 
advance of a Hellenistically Oriental culture and the pathetic attempts to resist it are 
among the most illuminating developments of the 2d and ist centuries b.c. (see Voigt, 
1893: passim). 

b. At this time the Roman statesmen began "to look for guidance not to Athens or 
Sparta but to the Persian Empire and the Hellenistic monarchies which succeeded it" 
(Stevenson, 1934: 183). 

CHAPTER 6, E 211 

cursus publicus, and very consistently, he combined it with an elabo- 
rate intelligence service. 23 

These steps were supplemented by such developments as the em- 
ployment of former slaves, "freedmen," in the service of the state, 2 * 
the use of eunuchs for political purposes, 25 the worship of the em- 
peror, and the gradual decay of independent commercial and indus- 
trial enterprise. Long before the close of the 2d century a.d., when 
Septimius Severus through wholesale slaughter and confiscation 
made the despotic center the "owner of most of the good arable 
land throughout the empire," 2e the old society had lost its identity. 
It was only logical that the "Semitic emperor," who despised Italy 
and "spoke Latin with a Punic accent," 2T wanted to be called do- 
minus, "master." c 

Thus when Diocletian established a spectacularly Eastern court, 
the actual Orientalization of the empire had already been accom- 
plished. A prominent economic historian summarizes the great trans- 
formation as follows: "In the second and third centuries . . . not 
only was the State (or the emperor) the largest landed proprietor, it 
was also the biggest owner of mines and quarries, and in course of 
time came to be the greatest industrialist." 28 Furthermore, "trade — 
wholesale and retail — became increasingly subject to governmental 
control" 29 and "transport was also largely nationalized." 80 In this 
single-centered economic setting, "the idea of the omnipotence of 
the State" evolved readily. It took shape essentially "under the influ- 
ence of orientalizing-hellenistic and other theories of the State." The 
wholesale "replacement of one economic system by the other, and 
the substitution of a new civilization and attitude to life for the old 
took more than a century and a half. It was completed by the end 
of the third century." 31 

A comparative analysis of the Orientalization of the Roman Em- 
pire leads to certain basic conclusions: 

1) The institutional meaning of this process appears clearly only 
if its study is based on the understanding of hydraulic society and 
agromanagerial (Oriental) despotism. 

2) Hellenization means Orientalization. The Hellenization of 
Rome started almost two hundred years before the establishment of 
the principate. 

c. "It was as if the spirit of ancient Assyria had taken possession of the palace to 
make the Empire subject to a bureaucracy which should be the executive of a divine 
authority transmitted through a dynastic succession. In such a system there would be 
no place for a Senate or for the principle of delegation by the State, and it was a sign 
that this notion of government now tended to prevail that the title dominus came 
to be generally applied to the emperor" (Miller, 1939: 35). 


3) As a societal type, imperial Rome must be equated not with 
the proto-industrial absolutisms of the West, but with the great agro- 
managerial absolutisms of the East, 


Different from the absolutist rulers of post-Medieval Europe, the 
Roman administrators of Spain, Gaul, Western Germany, and Eng- 
land were not restricted by nationally organized property-based 
corporations (estates). And although they preserved as far as possible 
the indigenous political leadership and culture, they operated the 
political apparatus in accordance with the great traditions of agro- 
managerial statecraft. As elsewhere, they created huge nonhydraulic 
constructions, primarily state roads and frontier walls. By means of 
their state post they monopolized quick communications. And they 
counted and taxed the inhabitants of the Western provinces in much 
the same way as they did in the East. 32 

No innate Iberian, Celtic, or Germanic urge for freedom kept the 
ancestors of modern Western Europe from accepting — at first under 
coercion but later as a matter of course — the yoke of a state which 
gave the nongovernmental forces of society little chance to participate 
in shaping their political and economic fate. Over several centuries 
Oriental despotism in its Hellenistic-Roman form spread into the 
woodlands of Germany, to the Atlantic shores of Spain and Gaul, and 
to the southern borders of Scotland. 

These Eastern institutions did not disappear when, in the 4th 
century, Western Rome, for all practical purposes, became inde- 
pendent of the hydraulic East. The despotic state, which had tolerated 
no strong and organized proprietary classes — although it did tolerate 
large property of all kinds — continued to reproduce itself even after 
its managerial and bureaucratic apparatus shrank. Indeed, until the 
end, the government of Western Rome insisted upon its absolutist 
position. Its last prominent political figure, Heraclius, was a typical 
representative of hydraulic statecraft, a eunuch. 33 

As in Late Byzantium, the decline of Western Rome was largely 
due to external factors. The loss of revenue from the wealthy eastern 
provinces seriously weakened the Italian metropolis, which was also 
having great difficulty in adjusting itself to the collapse of its slave 
economy. The East, being agriculturally more intensive, never had 
relied on slave labor as had the West. And consequently the West 
suffered severely when the sources of cheap slave labor dried up. 

CHAPTER 6, E 213 

The political impotence of Rome became blatantly apparent at 
the beginning of the 5th century: Rome lost Gaul in 406, England in 
407, Spain in 415, and Africa in 429. Within the truncated metropolis, 
the forces of big landed property, as represented by a new senatorial 
group, increased in importance. However, the emerging proprietary 
leaders lacked the strength to set up a non-Oriental type of govern- 
ment. This objective was achieved only when they joined the Ger- 
manic king, Odovacar, who in 476 formally terminated the worn-out 
absolutism of Western Rome. 34 

c. Europe after 476 


Certain symbols of hydraulic statecraft, such as the vassals' obliga- 
tion to kiss the sovereign's foot, persisted for a considerable time, 
even outlasting the Merovingian period; 35 but lacking substantial 
societal foundations, they eventually ceased to be invoked. And the 
political development, instead of following the Roman model, pro- 
duced the decentralized protofeudal system of government which 
characterized the first period of the Middle Ages. 



In this period, which is assumed to have lasted until the end of the 
12 th century, 37 there appeared in 1086 the Domesday Book, a register 
of the lands of England, which was ordered in 1085 by the Norman 
king, William the Conqueror. European historians have indicated 
institutional roots of the Domesday both in England 38 and Nor- 
mandy. 39 But while these roots are entirely authentic, they do not 
adequately explain the great English-Norman land register. Not 
only was this type of public cadaster unknown in the area from which 
William and his men came ("Normandy had no Domesday and no 
dooms"), 40 but it was also unknown in other parts of non-Oriental 
Europe. According to Maitland, it represents "an exploit which has 
no parallel in the history of Europe." 41 

What then inspired this unparalleled achievement? Conquest, 
which Maitland suggests, 42 provides no plausible explanation, since 
Medieval Europe saw many conquests but only one Domesday Book. 
The Normans of Normandy are a case in point. They did not, to 
our knowledge, institute a Domesday, but they certainly settled in 
the north of France through conquest. Could it be that by 1085 


the Normans had become familiar with administrative methods 
which were unknown to them in the 10th or even in the earlier 
part of the nth century? 

When in 1066 the Normans conquered England, some of their 
countrymen had already set themselves up as the masters of southern 
Italy, an area which, with interruptions, had been under Byzantine 
administration until this date; and some of them had established a 
foothold in Sicily, an area which had been ruled by Byzantium for 
three hundred years and after that by the Saracens, who combined 
Arab and Byzantine techniques of absolutist government. 

We have no conclusive evidence regarding the effect of this By- 
zantine-Saracen experience on William and his councilors. But we 
know that in 1072 — that is, thirteen years before William ordered 
the descriptio of England — the Normans had conquered the capital 
of Sicily, Palermo, and the northern half of the island. And we also 
know that there were considerable "comings and goings" 43 between 
the Italian-Sicilian Normans and their cousins in Normandy and 
England, particularly among the nobility and clergy. The latter hap- 
pened also to be actively engaged in administrative work. 44 No won- 
der, then, that on the basis of his knowledge of the period Haskins, 
the leading English expert on English-Sicilian relations in the Middle 
Ages, suggests "the possibility of a connexion between Domesday 
Book and the fiscal registers which the south had inherited from its 
Byzantine and Saracen rulers." 45 

Haskins' hypothesis explains well why a typically hydraulic device 
of fiscal administration appeared in feudal Europe. It also explains 
why for hundreds of years afterward this "magnificent exploit" had 
no parallel in that area. Evidently, systematic and nationwide registra- 
tion was as out of place in feudal society as it was customary in the 
realm of Oriental despotism. 

d. Spain 


But neither the failure of the Frankish attempts nor the singularity 
of the English Domesday implies that after 476 the institutional 
divide between the hydraulic and nonhydraulic parts of Europe re- 
mained fixed. The history of southern Italy and Sicily prior to the 
Normans reveals two major forces of Eastern expansion: the By- 
zantines, who tried to uphold their way of government in certain 
former provinces of the Roman empire, and, far more significantly, 
the Arabs, who, inspired by a dynamic new creed and equipped with 

CHAPTER 6, E 215 

new methods of warfare, 46 extended their power from the Near Eastern 
centers of hydraulic society throughout Northwest Africa, Spain, and 
— temporarily — Sicily. 

This colossal eruption resembled the westward growth of the 
Roman empire in that it, too, spread Orientally despotic patterns of 
government. But for a variety of reasons the institutional effects of 
the Islamic conquest were much more far-reaching. Under Roman 
influence Western Europe became part of a loosely hydraulic Oriental 
society without, however, adopting hydraulic agriculture; and eventu- 
ally it returned to a submarginally hydraulic or altogether non- 
hydraulic position. Under the influence of the Arabs, the swing was 
considerably greater. Prior to the Islamic invasion, the Iberian 
peninsula was the home of a protofeudal civilization, which had 
small-scale irrigation agriculture but probably few hydraulic enter- 
prises/ In sharp contrast to the Romans, who seized Western Europe, 
the Arab conquerors of Spain were entirely familiar with hydraulic 
agriculture, and in their new habitat they eagerly employed devices 
that had been extremely profitable in the countries of their origin. 
Under Muslim rule "artificial irrigation . . . was improved and ex- 
tended ... on Oriental models," and this included government 
management: "its superintendence was the business of the state." " 

Thus Moorish Spain became more than marginally Oriental. It 
became a genuine hydraulic society, ruled despotically by appointed 
officials 48 and taxed by agromanagerial methods of acquisition. The 
Moorish army, which soon changed from a tribal to a "mercenary" 
body, 49 was as definitely the tool of the state as were its counterparts 
in the Umayyad and Abbassid caliphates. A protoscientific system 
of irrigation and gardening 50 was supplemented by an extraordinary 
advance in the typically hydraulic sciences of astronomy and mathe- 
matics. 51 Contemporary feudal Europe could boast of no comparable 
development. Reconstructing the impressions of the great Arab geog- 
rapher, Ibn Hauqal, who visited Spain in the 10th century, Dozy 
comments on the organizational power of the Muslim state, whose 
police, like its hydraulic agriculture, penetrated the most remote 
parts of the country: "The foreigner noticed with admiration the 
universally well cultivated fields and a hydraulic system, which was 
coordinated in such a profoundly scientific manner that it created 
fertility in the seemingly least rewarding soils. He marvelled at the 
perfect order that, thanks to the vigilant police, reigned even in 
the least accessible districts." 52 

d. Hirth, 1928: 57 ff.; Hall, 1886: 363, 365; Levi-Provencal, 1932: 166; Laborde, 1808: 
29, 107. Laborde's memoir claims complete lack of agricultural interest for the Gothic 
conquerors of Spain (Laborde, 1808: 107). 


In the second half of the 14th century the leading town of the 
Hanseatic League, Liibeck, numbered 22,000 inhabitants, 53 and Lon- 
don about 35,ooo. 54 At the height of the Western caliphate the 
Moorish capital, Cordoba, may have harbored a million persons, 55 
and Seville in 1 248 more than 30o,ooo. 56 At the close of the Muslim 
period Granada was probably at least as populous. The Encyclopedia 
of Islam estimates the dwellers of this beautiful last Islamic capital 
in Spain at "half a million." 57 

No wonder then that the absolutist state, at the peak of its pros- 
perity, collected a stupendous revenue. 58 And no wonder either that 
this state, which like other hydraulic regimes freely used eunuchs, 69 
was ruthless in purging dignitaries who fell from favor. When these 
unfortunates were liquidated, the state was quick to confiscate what- 
ever property they possessed 



The reconquista, which in the 13th century reestablished Chris- 
tian control over the greater part of Spain, transformed a great 
hydraulic civilization into a late feudal society. Students of Russia, 
who see the rise of an Orientally despotic state in Muscovy as the 
consequence of an armed struggle against powerful Eastern enemies, 
will do well to compare the Russian story with what happened in 
Spain — and for that matter, in Austria. 

To begin with the latter country. For several centuries, Austria 
was threatened by one of the greatest Oriental empires known to 
history: Ottoman Turkey; and extended parts of Hungary were oc- 
cupied by the Turks for more than one hundred and fifty years. 
But the main political and military base of the counterattack, Austria, 
remained free; and the protracted struggle against the mighty Eastern 
foe did not convert the Austrian state into an Oriental despotism. 
Like other countries of Europe, Austria advanced toward a definitely 
Western type of absolutism: until the middle of the 18th century the 
Austrian diets (Landtage) had a decisive voice concerning taxation 
and the drafting of soldiers, 61 and even after 1740 the estates played 
an essential role in the fiscal administration. 62 Hungary stubbornly 
maintained a semi-autonomous government, whose Landtag, con- 
sisting of an upper house (clerical and secular magnates) and a lower 
house (lower nobles and urban deputies), "exerted a great influence 
on the country's administration." 63 

In Spain, too, the base of the Reconquest was never Orientalized. 
The rulers of the small northern states that had withstood the Arab 
onslaught depended for their military strength on the support of the 

CHAPTER 6, E 217 

nobles, the clergy, and the towns; 64 and at the end of the main phase 
of the Reconquest these groups, far from being politically pulverized, 
were able, because of their privileges, to maintain a semi-autonomous 
existence. 65 Similar to the development in late feudal and postfeudal 
France, England, Germany, Italy, and Scandinavia, Spain also de- 
veloped an absolutist government. 66 This government was strong 
enough to prevail over the nobles, the Church, and the towns; 6T but 
it was unable to wipe out the entailed aristocratic landholdings 68 and 
the semi-autonomy of the Church; and it was unable to break the 
pride and dignity of the Spanish people. The estates of Aragon that 
had declared the recognition of their privileges to be the condition 
of their homage to the king ("si no, no.") repeated this daring formula 
again in 1462, 69 that is, more than a hundred years after the greater 
part of the Peninsula had been reconquered. And although the as- 
semblies (cortes), which in Castille essentially represented the free 
municipalities, had ceased to exist in 1665, the absolutist regime failed 
to instill in its subjects the submissive attitude habitual under hy- 
draulic regimes. 

To state this is not to deny the extraordinary strength of Spanish 
absolutism. This phenomenon may at least in part be explained by 
the exigencies of the Reconquest "frontier," which enhanced the 
growth of royal authority in Catalonia, Navarre, and Aragon. 70 How- 
ever, the Wirtschaftsgesinnung of the Christian kings may have been 
even more decisive. The northern base of the Reconquest greatly 
favored pastoralism; and the European demand for wool — which in- 
creased with the advance of the Reconquest 71 — led the Spanish kings 
one-sidedly to promote sheep breeding also in the liberated areas 
of central Spain, and even in parts of southern Spain. 72 While the 
kings gave all manner of privileges to the towns and nobles, they 
established a tight fiscal and jurisdictional control over the sheep 
breeders, who, from the 13 th century on, were combined in a special 
organization, the Mesta. 73 

In Spain, as in England, the sheep "ate" the people. But Spain 
differed from England in that, almost from the beginning, the princes 
profited enormously from the rapidly expanding pastoral economy. 
State revenues from this source were large. 74 Eventually the monarchs 
considered "the exploitation and conservation of the pastoral in- 
dustry . . . the principal sustenance of these kingdoms." 75 

The huge revenues which the Crown received from its colonial 
empire have frequently been held responsible for the decline of the 
Spanish population in the 16th century/ However, the depopulation 

e. Seville, which in 1247 ^ a< ^ over 3°°» 000 inhabitants, numbered in the 16th century 
200,000. Cordoba, which under the caliphs may have harbored a million people, now 


of the villages, which certainly was a major cause for the depopula- 
tion of the cities, cannot be satisfactorily explained thus, since the 
influx of gold and silver would have enabled the enriched towns- 
people to buy more rather than fewer rural products. 

In all probability the downward trend was caused primarily by the 
replacement of labor-intensive irrigation farming by labor-extensive 
cattle breeding. This development, stimulated by the soaring export 
of wool/ led to the promulgation of the Leyes de Toro, which com- 
pleted the "subjection of agriculture to large scale pasturage" 76 four- 
teen years before Cortez took Mexico and twenty-eight years before 
Pizarro took Cuzco. And it also accounts for the great reduction in 
the Indian farming population in post-Conquest Mexico, Yucatan, 
and Peru." 

In the Spanish countryside, herds and herders now made their 

numbered 60,000 (Laborde, 1808: 9). The population of Granada decreased from per- 
haps 500,000 to 80,000 (see above, and Laborde, 1808: 9). These decreases resulted in 
part from military destruction; but in part they express the transformation of the 
rural order. Some sections of the countryside never recovered from the pestilence 
and the Reconquest (Klein, 1920: 337). Others were allowed to lie fallow during the 
16th and 17th centuries (ibid.: 320, 342 ft.), until the formerly flourishing fields were 
"smitten with the curse of barrenness" (Prescott, 1838, III: 461, n. 85), because sheep 
breeding had been allowed "to run riot throughout the land and to annihilate almost 
the last vestiges of agriculture that still remained" (Klein, 1920: 343). 

The known ruins of former settlements in Catalonia, Aragon, Leon, Valencia, Mancha, 
Castille, etc. numbered more than 1141. The region of the Guadalquivir had boasted 
1200 villages under the Caliph of Cordoba. In 1800 only 200 survived. Of the fifty 
villages of Malaga, only sixteen were left. One section of the diocese of Salamanca had 
only 333 villages out of its former 748, while of 127 villages which existed near des 
partidos de Banos pena del rey only thirteen remained (Laborde, 1808: 8). The area 
of the kingdom of Granada that prior to 1492 had supported three million people 
numbered only 661,000 by 1800 (ibid.: 9). 

/. The rise continued until the latter part of the 16th century (Klein, 1920: 37-46). 

g. Ships were small and freight expensive; and nothing much was to be gained 
by exporting grain to Europe. Silver was the most highly prized export article; but 
handsome profits could also be made in sugar and cacao, dyewoods, dyestuffs, and hides 
(Humboldt, 1811, IV: 368 ff.). Within a few decades "oxen, horses, sheep, and pigs 
multiplied to a surprising degree in all parts of New Spain" (ibid., Ill: 224). By 1570, 
when Acosta arrived in America, some individuals owned as many as 70,000 and even 
100,000 sheep (Acosta, 1894, I: 418; Obregon, 1928: 151). Wherever the increase of 
cattle was not checked, the herds grew rapidly, not only in Central America but 
also in the Southwest of North America (Obregon, 1928: 151), in Peru (cf. Markham, 
1892: 163; see also Juan and Ulloa, 1806, I: 300, 318, and passim), and in Yucatan 
(Shattuck, Redfield, and MacKay, 1933: 15). When Cortez set up a princely estate in 
Oaxaca, he at once "imported large numbers of merino sheep and other cattle, which 
found abundant pastures in the country around Tehuantepec" (Prescott, 1936: 671). 
Consistently, it was Cortez who in the New World organized a Mesta patterned after 
the Mesta of Castille (Mendoza, 1854: 225). 

CHAPTER 6, E 219 

lonely way over vast grasslands. It was in this landscape that Don 
Quixote urged on his stumbling nag. And in the cities no spectacle 
was so popular as the bullfight. In Valladolid, in 1527, Charles V 
celebrated the birth of his son, Philip II, by himself entering the 
ring to challenge the bull. 

e. The Introduction of Oriental Despotism into Russia 

"The Tatars had nothing in common with the Moors. When they 
conquered Russia, they gave her neither algebra nor Aristotle." 
Pushkin was doubtless correct in lamenting the negative cultural 
consequence of the Tatar h conquest. He might have gone even 
further and noted the devastating political consequences of their 
fabulous military success. The Tatars, who by 1240 had crushingly 
defeated the Eastern Slavs, controlled their new subjects so effectively 
that no independent Russian power undertook to liberate them. 

Nor did any internal Russian force engage in a systematic and 
open struggle against the Horde. The isolated military victory at 
the Don River, which the Grand Duke of Moscow, Dmitry, won 
over a Tatar army in 1380, backfired sadly: the subsequent reprisals 
discouraged armed resistance for another hundred years. 1 Even when, 
in 1480, Ivan III refused allegiance to the enfeebled Tatars, he 
avoided battling against them. The Tatars, while still able to lead 
an army against the Muscovite host, were equally reluctant. Inde- 
cision on both sides resulted in "an unbelievable spectacle: two 
armies fleeing from each other without being pursued by anyone." 
To quote Karamsin further: "So ended this last invasion of the 
Tatars." > 

So indeed ended Tatar rule over Russia. It had lasted for almost 
two hundred and fifty years; and the Grand Duchy of Muscovy, which 
rose to prominence during this period, did so not as an independent 
force but as the instrument of the Khan. 

h. The name "Tatar" originally referred to peoples living in the eastern part of 
Inner Asia (see Wittfogel and Feng, 1949: 101 ff.). After the great expansion of Mongol 
power during the 13th century, the name began, in Eastern Europe, to denote those 
Mongols and Turks who together formed the core of the Golden Horde. Merging 
with older Turkish and Finnish groups, these "Tatars" spoke Turkish, a language 
which by then had become the most important ethnic and cultural trait of the 
westernmost sector of the Mongol world (Spuler, 1943: 1 1 n.). In the present discussion 
the terms "Tatar" and "Mongol" are used interchangeably to designate the people 
of the Golden Horde. 

1. After 1380 the leading principality, Muscovy, "for the time being did not think 
of fighting the Tatars" (Kliuchevskii, Kurs, II: so). 

j. Karamsin, HER, VI: 195-6. 


This fact is not disputed. Nor is it seriously denied that 16th- 
century Muscovy cannot be equated with Western absolutism. How- 
ever, opinions differ fundamentally concerning the origin of Mus- 
covy's despotism. Was Ivan's autocratic control over land and people 
due to external conditions, namely to a continually fought-over 
frontier? Or was it due primarily to the influence of the Mongols 
who in Russia applied despotic methods of statecraft learned in sev- 
eral hydraulic countries of Asia, particularly China? /M * 

Historians who uphold the "external" interpretation lean heavily 
on the authority of the foremost modern Russian historian, Rliu- 
chevsky. I fully share the esteem in which he is held by scholars of the 
most diverse opinion; but I find his views on the emergence of Mus- 
covite despotism less one-sided than is generally assumed. 

True, Kliuchevsky has paid little attention to the Tatar Yoke,* 
and his understanding of Oriental despotism is limited." 1 But he 

j-bis (added to Third Printing). The Mongols were familiar with the organizational 
and acquisitive methods of Chinese statecraft when they subdued Russia (1237-40). 
Chingis Khan conquered China north of the Yellow River (1211-22) and Turkestan 
(1219-20). From 1215 on he had as a top-ranking adviser a Liao-Chinese, Yeh-lii Ch'u- 
ts'ai (Wittfogel and Feng 1949: 669 f.), who later also served his son Ogotai (1228-41). 
Ogotai completed the conquest of North China in 1234. By 1240 the Mongols had 
learned to operate a state post, and they managed in North China census- taking, taxa- 
tion, and the corvee (Yuan Shih 2: lb, 2a, 7a; 121: 9a; 146 passim; 191: 2a. Cf. also 
Hsin Yuan Shih, chap. 127; Chingis Khan's Yasa and The Secret History of the 
Mongols). In the 1240's Carpini observed the state post and the taking of a preliminary 
census in Mongol-dominated Russia. In 1253 the Great Khan, Mongke, ordered a certain 
Pieh-erh-ke (Berke?) to take a census in Russia (Yuan Shih 3: 4b). Russian sources report 
that this was done in 1257; and they mention for 1259 a Mongol census-taker, "Berkai" 
(Karamsin, HER, IV: 91, 94; E. Bretschneider, Mediaeval Researches, London, 1910. II: 
80). The Great Khan's jurisdiction over the Golden Horde lasted until 1259 (Spuler 
1943: 41 f. and 252), that is, throughout the formative years of the Tatar Yoke. 

k. Florinsky criticizes him for suggesting that when studying the political organiza- 
tion of northeastern Russia, one "should forget for a time . . . that Russia was con- 
quered by the Tartars" (Florinsky, 1953, I: 78); and Vernadsky (1953: 333 ff.) notes 
that except for "a few general remarks on the importance of the khans' policies for 
the unification of Russia . . . [Kliuchevsky] paid little attention to the Mongols." 

m. Kliuchevsky was not too familiar with the institutions of Oriental society and 
with such of its variants as traditional China. Otherwise he would not have contrasted 
the service-based class system of Muscovite Russia and the conditions of Oriental 
despotism (Kluchevsky, HR, III: 52). In another context, however, he notes the 
similarities in the Muscovite methods of liquidating potentially dangerous relatives 
and methods of Oriental despotism in like situations (ibid., II: 88). And his description 
of state service and land tenure in post-Mongol Russia clearly indicates institutional 
affinities to Ottoman Turkey and Muslim India. His discussion of Peter's efforts to de- 
velop industry is a major contribution to our understanding of the Russian version of 
an agrobureaucratic despotism. The omnipotent state, based on enforced service and 
claiming ultimate control over all land, has also been viewed as a key element of 

CHAPTER 6, E 221 

was too great a student to overlook the crucial institutional changes 
which,' under Tatar rule and because of it, occurred in Russia's state 
and society. According to his own account, these changes definitely 
preceded the rise of the "frontier," with whose formative role he 
is so impressed. 

Indeed, Kliuchevsky, in his "frontier" thesis, deals essentially with 
the post-Tatar period. He describes the changes involved in the 
recruiting of "a numerous military-official class" as being closely 
connected with "the territorial expansion of the empire," whose new 
frontiers had "placed the state in direct contact with such external 
and alien foes of Russia as the Swedes, the Lithuanians, the Poles, 
and Tatars. This direct contact had put the state in such a position 
that it had come to resemble an armed camp surrounded on three 
sides by enemies." 77 Manifestly, the Tatars of whom Kliuchevsky is 
speaking are those that confronted 16th-century Muscovy, and the 
frontier of which they form a part is the 1 6th-century frontier. Kliu- 
chevsky says so expressly, 78 and several times he refers specifically to 
the years from 1492 to 1595.™ 

In view of these facts we cannot help feeling that Kliuchevsky's 
"frontier" thesis raises more questions than it answers. Why should 
a non-Oriental Russia evolve into an en forced-service despotism, be- 
cause Russia was fighting such Western countries as Sweden, Lith- 
uania, and Poland? Many European governments dealt with compa- 
rable enemies without establishing Orientally despotic patterns of 
control over land and people. Or why should a non-Oriental Russia 
become Orientally despotic when the Oriental forces she was com- 
bating were, relatively speaking, no stronger than the Turks, with 
whom the Austrians and Hungarians fought, or the Moors, with 
whom the Spanish reconquerors were engaged in a life and death 
struggle? Neither Hungary and Austria nor Spain became Orientally 
despotic because of their Oriental "frontier." We may therefore 
well ask: Could the Muscovite development of the 16th century have 
occurred because Russia, prior to this period and as the result of 
her long subjection to Oriental domination, had already taken de- 
cisive organizational and acquisitive steps in the direction of a des- 
potic "service" state? 

Kliuchevsky's frame of reference prevents him from giving a con- 
sistent answer to these questions. But it is amazing how far his account 

Tsarist society by Sumner, who considers Tsarism to be rooted in the "ideas and 
ritual" of Byzantium and "the fact and practice of the Tatar khans." Elaborating this 
point, Sumner observes that it was under the influence of the Golden Horde rather 
than "far-away Byzantine administration" that the Muscovite government and military 
system originated (Sumner, 1949: 82 f.). 


of 13th-, 14th-, and 15th-century Russia goes in affirming the socio- 
historical significance of the Tatar period. 

According to Kliuchevsky, it was during this period that the 
towns, which had played a prominent role in Kievan Russia, 80 lost, 
with a few exceptions (Novgorod, Pskov), their political impor- 
tance; 81 and it was in this period that the territorial princes and in- 
dependent boyars, after a temporary improvement in their conditions, 
were sharply curbed by the grand dukes of Muscovy. Many princes 
became the serving men of Muscovy, whose new prince-officials by 
1500 "overlaid, if they did not crush, the older stratum of Muscovite 
non-titled boyars." 82 

Why did this happen? In the matter of the political emasculation 
of the towns, Kliuchevsky shuts his eyes to the effects of Tatar rule," 
which were pointed out earlier by Karamsin. In the matter of the 
fate of the boyars and territorial princes, he recognizes that Tatar 
power enabled Muscovy to subdue them. 

Kliuchevsky is aware that for more than two generations the Tatars 
operated the fiscal organization that they had erected in Russia: 
"After their conquest of Rus, the Tartars themselves first collected 
the tribute they imposed on Rus." 8S He is also aware that political 
and jurisdictional power accrued to Moscow when, in 1328, the 
Khan transferred this function to his Muscovite deputy: "The simple 
trustee-agent in charge of collecting and delivering the tribute of 
the Khan, the Prince of Moscow, was then made plenipotentiary 
leader and judge of the Russian princes." Subsequently the Khan's 
commission became "a powerful instrument for the political unifica- 
tion of the territorial states of Rus." 8 * 

In all these instances, Tatar influence is clear. It becomes still 
more impressive when we recognize the bureaucratic innovations that 
accompanied the political change. Kliuchevsky knows that the meth- 
ods of registering land and tax payers which were used throughout 

n. Kliuchevsky views this development as the result of the colonization of northern 
Russia (Kluchevsky, HR, I: 269). "Rus" did indeed expand northward, but this is only 
half the story. In Western Europe many towns, which were founded by princes or 
feudal lords, emancipated themselves. Why was it that in 13th- and 14th-century 
Russia princely authority grew at the expense of the towns? And why did the veche 
cease to function, even where it had previously prevailed? 

o. Karamsin (HER, V: 451) ascribes the change to the increased authority with 
which the princes were endowed by the Tatars. Recently Vernadsky noted that "the 
destruction of most of the major cities of East Russia during the Mongol invasion" 
was followed by an equally devastating, and even more successful, political campaign 
against the towns and that in this campaign the Russian princes and boyars supported 
their Mongol masters. In the middle 14th century, the veche "had ceased to function 
normally in most East Russian cities and could be discounted as an element of 
government" (Vernadsky, 195$: 345). 

CHAPTER 6, E 223 

the 16th and 17th centuries 85 had existed at the close of the 15th 
century and long before." He knows that after the conquest of Russia 
the Tatars "during the first thirty-five years of the Yoke three times 
took a census, chislo, of the entire Russia people, with the exception 
of the clergy, by means of chislenniki [census takers] sent from the 
Horde." 86 Subsequent studies have thrown additional light on the 
original Tatar organization, 87 which may have served military as well 
as fiscal purposes. 88 Vernadsky plausibly suggests that "it was on the 
basis of the Mongol patterns that the grand ducal system of taxation 
and army organization was developed in the late 14th to 16th cen- 
turies." 89 His conclusion elaborates what Kliuchevsky had intimated 
fifty years before. 

When describing the state post of 16th-century Moscow, 90 Kliu- 
chevsky does not expressly connect it with earlier developments. But 
his remark, "the Jamskoi prikaz, the Department of Posts, which 
was known from the beginning of the 16th century," 91 in all likeli- 
hood points to Ivan III, 92 that is, to the close of the Tatar period. 
Other scholars have connected the postal system, yam, which the 
Tatars maintained in Russia, 93 with the Muscovite institution of the 
same name. 9 

The rise of Muscovite despotism coincides with the rise of the new 
type of civil and military serving men, who, as temporary holders of 
state land (pomestye), were unconditionally and unlimitedly at the 
disposal of their supreme lord. From the later part of the 14th century 
on, the grand dukes of Muscovy began to reduce the territorial 
princes to the position of serving men; 94 and in the 15th century 
they assigned office land — which was previously given only to unfree 
retainers 95 — to free serving men as well, mainly to warriors but 
also to civil ("court") officials. 96 Kliuchevsky is fully cognizant that 
this type of compulsory service differs from the conditions of Western 
Europe; 97 and it is therefore not surprising that in his discussion 
of the legal principles involved in the institution of the pomestye 

p. Kluchevsky, HR, III: 228. The Tatar origin of the Muscovite system of census 
taking has been stressed among others by Miljukov (1898: 128) and Kulischer (1925: 
404), the latter of whom, not without reason, assumes ultimate Chinese influence. 

q. Bruckner, 1896: 521 ff.; Milukow, 1898: 81; Kulischer, 1925: 405; Grekov, 1939: 
216 ff. The Altaic term yam, "post," and jamti, "postmaster" (Spuler, 1943: 412) 
appeared in Russian as jam and jamshchik (Briickner, 1896: 503, 522). During the 
Mongol period, "the yam was a special tax for the upkeep of post-horse stations" 
(Vernadsky, 1953: 221). When in the early part of the 16th century Herberstein used 
the Muscovite state post, he had relay horses assigned to him "by the post-master, who 
in their language is called 'jamschnik* [sic!]." The relay stations were called "jama" 
(Herberstein, NR, I: 108). In the 16th century the Postal Chancellery was first called 
jamskaja izba, then jamskoj prikaz (Staden, 1930: 13, n. 4; cf. 15, 59). 


he considers only two roots, both Oriental: Byzantium and the Tatar 
Horde. Rejecting the former, he is left with the Tatar alternative, 
suggested by Gradovski. According to this view, "the idea of the 
prince as the supreme landowner originated only during the Mongol 
period. As representatives of the authority of the Khan, the Russian 
princes enjoyed, in their territories, the same rights as the Khan 
himself enjoyed in all the territory under his rule. Later the Russian 
princes inherited these state rights from the Khan completely; and 
this shattered the incipient private ownership of land." 98 

It is characteristic for Kliuchevsky's ambivalence toward the Tatar 
issue that he fails to verbalize what, from the standpoint of his own 
premises, is the only logical conclusion. But he does not hesitate to 
stress the rapid growth of the pomestye institution at the close of the 
Tatar period. Evidently, "traces of an intensive and systematic dis- 
tribution of public land in pomestye tenure can already be found 
during the second half of the 15th century." " The Muscovite princes 
established pomestye lands on a large scale first in newly conquered 
territories such as Novgorod; but in the early 16th century "a great 
development of pomestye tenure" also took place in the vicinity of 
Moscow. 100 

The comparative economic historian, Kovalevsky, expressly claimed 
a Tatar origin for the fateful institution: "It is a fact that prior to 
the 15th century we never hear of Russian princes paying for services 
except by the distribution of money and objects taken as war loot, 
whereas the assignment of military tenures under the name of iktaa 
was known in the entire Mohammedan world and especially among 
the Tatars for centuries prior to the appearance of this practice in 
Muscovy. These considerations led the author to state that this kind 
of practice was introduced to Muscovy and the other Russian prin- 
cipalities through the imitation of the Tatar khanates." 101 Vernadsky 
does not claim a direct link; but he too calls the Mongol age the 
"incubation period" of the pomestye system. 102 

In view of these facts it is hard to reject Vernadsky's conclusion 
that in the days of the Tatars the old free society of Kievan Russia 
was "persistently chipped away without at first affecting the facade," 
and that when Ivan III broke with the Horde, "the framework of 
the new structure was all but ready and the new order, that of a 
service-bound society, became clearly noticeable." 103 

It became clearly noticeable indeed. And a few decades after Ivan's 
death, the forces of despotism had gained sufficient strength to destroy 
ruthlessly the obsolete facade. The time lag between incubation and 
maturation reflects the contradictory interests of the Tatars, who 
wanted their Muscovite agency to be sufficiently strong to carry 

CHAPTER 6, F 225 

out the will of the Khan but not strong enough to override it. With- 
out foreseeing the ultimate consequences of their action, they built 
an institutional time bomb/ which remained under control during 
their rule but which started to explode when the "Yoke" collapsed. 
Byzantium's influence on Kievan Russia was great, but it was 
primarily cultural. Like China's influence on Japan, it did not seri- 
ously alter the conditions of power, class, and property. Ottoman 
Turkey's influence on 16th-century Russia stimulated a regime that 
was already Orientally despotic, 104 but it did not bring it into being. 
Tatar rule alone among the three major Oriental influences affecting 
Russia was decisive both in destroying the non-Oriental Kievan 
society and in laying the foundations for the despotic state of Mus- 
covite and post-Muscovite Russia. 


Thus Greece, Rome, Spain, and Russia all crossed the institutional 
divide. In Greece, Rome, and Spain the pendulum swung back and 
forth. In Tsarist Russia the reverse movement (away from a despotic 
state) came close to bringing the country back into the Western orbit. 
The changes that occurred in each of these cases were enormous; 
but their character cannot be clearly understood unless the affected 
institutional structures are clearly defined. Our analysis has tried 
to do this. Approaching both structure and change from the stand- 
point of varying hydraulic and bureaucratic density, we can draw 
the following major conclusions. 

1. Structure 

a. Density Subtypes of Hydraulic Society 

There are two subtypes of hydraulically compact areas: one with 
economically predominant and continuous hydraulic systems (Com- 
pact 1), the other with economically predominant but discontinuous 
hydraulic systems (Compact 2). There are two subtypes of hydrau- 
lically loose areas: one with an organizationally predominant hy- 
draulic system, which comprises major and regionally compact 
hydraulic units (Loose 1), the other without major compact units 
(Loose 2). And there are two subtypes of the margin of hydraulic 
society: one containing conspicuous hydraulic elements (Margin 1), 
the other lacking such elements (Margin 2). A seventh subtype, the 
submargin, belongs to the fringe of the hydraulic world, because its 
r. Vernadsky (1953: 335) appropriately speaks of "influence through delayed action." 


representatives employ conspicuous elements of Orientally despotic 
statecraft. But since its dominant institutions are of a definitely non- 
hydraulic character, it must be placed on the outer fringe of this 

b. Differing Frequencies of Occurrence 

The hydraulically densest subtypes of hydraulic society, Compact 1 
and 2, are not the most frequent ones. Nor can the other subtypes 
be called less "advanced," if this term is meant to imply that eventu- 
ally and necessarily they will become compact. Among the historically 
prominent hydraulic societies, and particularly among their larger 
representatives, the compact patterns are more the exception than 
the rule. 

c. Decreasing Importance of Hydraulic Economy Proper 

The decreasing importance of hydraulic economy proper becomes 
clearly apparent when the agromanagerial world is viewed in its 
spatial and temporal entirety. There is little doubt that representa- 
tives of this world had a greater hydraulic density during their forma- 
tive and primary phase than during their later and secondary develop- 

In the formative phase, relatively small hydraulic commonwealths 
arose in semi-arid and arid settings. And if our genetic hypothesis is 
correct, we are safe in assuming that while, during this phase, a 
number of marginal hydraulic societies originated through diffusion, 
few such societies originated through the disintegration of larger, 
loosely hydraulic units, which were then practically nonexistent. The 
greatest number of marginal hydraulic societies — both absolutely 
and in proportion to the number of hydraulic societies proper — ap- 
peared therefore not during the formative phase, but after it. 

This developmental peculiarity is accompanied by another which, 
although independent of it, aggravates its effects. For reasons which 
in the Old World are closely connected with the spread of nomadic 
conquest and globally with the lessening of hydraulic concern, hy- 
draulic societies proper tend to reduce rather than increase their 
hydraulic intensity. 

The specific density patterns of industrial and hydraulic society 
develop in different ways. The representatives of industrial society 
tend to become more industrial without of necessity becoming in- 
dustrially compact. Conversely, the representatives of agromanagerial 
society seem to reach their highest hydraulic density coefficient during 
a relatively early phase of their growth. Afterward they hold their 

CHAPTER 6, F 227 

own or recede. Taken in its entirety, agromanagerial society ap- 
parently "advances" not to higher but to lower levels of hydraulic 

2. Capacity for Societal Change 

Our density analysis clarifies both structure and change. And it 
clarifies change — or lack of change — not only within the same societal 
type, but also from one societal type to another. 

1) The formation of hydraulic society apparently depends on the 
presence of a hydraulic economy proper as an essential condition. 

2) The perpetuation of hydraulic society is assured by a plurality 
of factors, among which hydraulic enterprise may be of little impor- 
tance, except in crises caused by the impact of strong external non- 
hydraulic forces. 

3) In a given hydraulic area, large government-controlled pro- 
ductive and protective water works may serve only a fraction of the 
politically dominated territory. The uneven diffusion of institutions 
of a given societal order, which characterizes the hydraulic world, also 
characterizes modern industrial society. Prior to World War II, the 
U. S. A. was an outstanding case of an industrial society. But, at that 
time, out of some 3,000 counties only some 200 — that is, about 7 per 
cent — were classed as "industrial counties" proper. 

4) The history of hydraulic society records innumerable rebellions 
and palace revolutions. But nowhere, to our knowledge, did internal 
forces succeed in transforming any single-centered agromanagerial 
society into a multicentered society of the Western type. 

5) More specifically: neither in the Old nor in the New World did 
any great hydraulic civilization proper spontaneously evolve into an 
industrial society, as did, under nonhydraulic conditions, the coun- 
tries of the post-Medieval West. In the marginal hydraulic civiliza- 
tion of Late Byzantium the rise of big private property led only to 
societal paralysis. In Russia, after severe attacks from the outside, the 
forces of private property (and their concomitant, free labor) pre- 
vailed in 1917 for a number of months over the system of despotic 
state power. 

a. For a detailed account of this phenomenon see The Structure of the American 
Economy, Pt. I, Basic Characteristics (Washington, D. C, National Resources Com- 
mittee, 1939), p. 47. 



atterns of proprietary complexity in 
hydraulic society 

Not all hydraulic societies comprise independent proprietary forces 
of consequence. When such forces are present, they seem to be more 
of a threat to the margin than to the hydraulic heartlands, although 
even in these latter, strong proprietary developments intensify social 
differentiations and periodic political crises. 

Hence, an institutional analysis of hydraulic society should deal 
not only with the density of its agromanagerial apparatus but also 
with the complexity of its proprietary development. Having explored 
the major patterns of hydraulic and bureaucratic density, we shall 
now examine the major complexity patterns of private property 
and enterprise which emerge under the shadow of agromanagerial 


Property is the individual's recognized right to dispose over a 
particular object. Like other rights, the right called property involves 
more than a relation between a person and a thing. It involves a 
relation between the proprietor and other individuals who, through 
the former's prerogative, are excluded from disposing over the object 
in question. 

The relation also involves the representatives of government, who, 
on the one hand, share the restrictions placed on the private non- 
proprietors and, on the other, are concerned with maintaining the 
existing property regulations. Thus in addition to being a legal and 
social institution, property is a political phenomenon. And property 
rights in different societies, even when they are similar in form, need 
not be similar in substance. 

Strong property develops in a societal order which is so balanced 
that the holders of property can dispose over "their" objects with a 
maximum of freedom. Weak property develops in a societal order 
that is not so balanced. 

a. In an incipient form this concept has already been used by Sir Henry Maine in his 
Village-Communities (New York, 1889), pp. 158 ft., 221 ff. 


CHAPTER 7, C 229 

The preceding chapters have described those peculiarities of hy- 
draulic society which, by making the state inordinately strong, tend 
to make private property inordinately weak. Of course, weakness is 
not nonexistence. Hydraulic society has given rise to many forms of 
private property that, as far as external appearance goes, have their 
parallels in other societies. Some of these forms show a different degree 
of development in different hydraulic civilizations, and these distinc- 
tions are so regular — and so manifest — that we can establish several 
subtypes of proprietary (and societal) complexity. 


The concepts of mobile and immobile property present obvious 
difficulties, but they have great advantages for our inquiry. Immobile 
property (essentially land) is the basis of private enterprise in the 
main branch of hydraulic economy: agriculture; and mobile property 
(tools, raw materials, merchandise, money) is the basis of its two 
most important secondary branches: industry (handicraft) and com- 
merce. Persons, too, may become the object of a proprietary relation. 
Like many other institutional conformations, hydraulic society also 
knows slavery. But unlike mobile and immobile property, slavery 
under agromanagerial despotism does not establish specific patterns 
of independent enterprise. We shall discuss the peculiarities of this 
type of slavery in the next chapter, which deals with classes. 


A holder of strong property may dispose over his property in a 
variety of ways. 

He may put his property to whatever use he wants, as long as he 
does not interfere with the rights of other members of the common- 
wealth. He may employ it actively, either in the economic sphere 
(for purposes of subsistence and material gain) or in the sphere of 
physical coercion (for purposes of promoting his and his group's 
material or political interest); or he may employ it passively, con- 
suming it for purposes of maintenance and pleasure. Occasionally 
he may decide not to use it at all. He may make a piece of wood into 
a bow to serve him in a hunt or raid, or into a digging implement 
for farming. He may employ a piece of land for raising what crops 
he wishes, or for grazing or hunting, or he may let it lie fallow. 

The holder of strong property whose active property produces gains 


because he, either alone or with, or through, others, employs it 
effectively, is free to enjoy these gains fully. He owns the calf as well 
as the cow. He is free to alienate his property at will. And he is free 
to determine who shall inherit it when he dies. 


1. Simple, Semicomplex, and Complex Patterns of 

The holder of weak property may enjoy only a shadow of these 
prerogatives, but this does not destroy his desire to act as freely 
as he can. He exercises his modest rights with respect to both mobile 
and immobile, passive and active property. In the sphere of mobile 
and active property they become institutionally important when 
the holders of such property employ it professionally and independ- 
ently in industry and commerce. Those who engage in handicraft 
or trade take a decisive step forward when they begin to devote 
themselves to these pursuits professionally, that is, full time. How- 
ever, such an advance effects no major societal change, as long as 
the professional craftsmen and traders constitute only a new sub- 
section within the class of government functionaries. It is only when 
they use their property to operate professionally and independently 
that they appear as a new class. The difference is not one of the "mode 
of production" — which may not change at all — but of the producers' 
and traders' political (and politically conditioned societal) position. 

Land is tilled professionally (that is, by peasants who spend most 
of their time farming) as soon as agriculture becomes an essential 
basis of subsistence. And elements of private (independent) land- 
ownership emerge relatively early. But the landowners, who often 
do not till their soil themselves, are in many Oriental societies pre- 
vented from expanding the sphere of private agrarian property, since 
most of the land is, in one way or another, regulated by the govern- 
ment. It is only when free (nonregulated) land becomes the dominant 
form of land tenure that private landownership becomes a societal 
phenomenon comparable to the predominance of independent pro- 
fessional handicraft and trade. 

Independent active property advances unevenly in its mobile and 
immobile sectors. These developmental differences are sufficiently 
clear and regular to permit distinction between at least three major 
patterns of proprietary complexity in hydraulic society: 

1) When independent active property plays a subordinate role 
in both its mobile and immobile forms, we are faced with a relatively 



simple pattern of property. We shall call this conformation a simple 
hydraulic society. 

2) When independent active property develops strongly in in- 
dustry and commerce but not in agriculture, we are faced with a 
semicomplex pattern of property. We shall call this conformation a 
semicomplex hydraulic society. 

3) When independent active property develops strongly in industry 
and commerce and also in agriculture, we are faced with the most 
complex pattern of property to be observed in hydraulic society. 
We shall call this conformation a complex hydraulic society. 

2. Supplementary Remarks 

a. "Simple I" and "Simple 11" 

How far can private and independent property advance in industry 
and commerce? And when does private ownership in land prevail over 
all other forms of land tenure? We shall attempt to answer both 
questions when we discuss the peculiarities of semicomplex and 
complex configurations of Oriental property. 

Another question, however, must be settled first. Are there, within 
hydraulic society, conditions under which professional representa- 
tives of industry and commerce are altogether absent or, for all prac- 
tical purposes, as good as absent? Such conditions do exist indeed. 
They occur essentially in hydraulic tribes, which for this and other 
reasons represent the most rudimentary variant of a simple hydraulic 
society. We distinguish the tribal type of simple hydraulic society, 
"Simple I," from the state-centered type of simple hydraulic society, 
"Simple II." 

Table 4. Patterns of Proprietary Complexity in Hydraulic Society (Schematized) 




Agriculture Pursued: 

Industry and Commerce Pursued: 








with Privately 
Owned Land 

In the Main ' On Basis of Private 
Professionally Property and in the 
Main ' Independently 

+ ' 

e - 
+ e 
+ + 


-J- Feature conspicuous 

1. The meaning of the qualification is explained in 
the text, p. 233. 

— Feature inconspicuous or absent 2. The circle © indicates a developmentally new 

). Farmer-craftsmen and producer- traders. 


b. Proprietary Complexity and Hydraulic Density 

Correlations between the patterns of proprietary complexity, 
on the one hand, and patterns of hydraulic density, on the other, 
are less easily established. The rise of property-based enterprise and 
social classes is due to several factors, among which hydraulic density 
is only one, and one that in a given area tends to change its quality 
very slowly and usually only because of changing relations with other 

This, however, does not imply the absence of significant correla- 
tions between hydraulic density and proprietary complexity. Of the 
two major evolutionary steps that hydraulic property may take, at 
least the first — the transition from a simple to a semicomplex pattern 
— may be greatly retarded, if not altogether blocked, when the under- 
lying agrarian order is hydraulically compact. Like the correlation 
between the rise of a state-centered simple hydraulic society and 
the advance of professional industry and commerce, this correlation 
will be clarified when we systematically discuss the characteristics of 
simple, semicomplex, and complex patterns of Oriental property. 


l. Nonspecific Aspects 

Agricultural tribes handle their property in many different ways; 
and this is as true for hydraulic as for nonhydraulic communities. 1 
In the simpler farming communities of Melanesia, South America, 
and Africa "movables are privately owned, but not land." 2 Similar 
trends are found also among important North American tribes; 
however, in Melanesia and West Africa a more differentiated pat- 
tern has made its appearance. "As a rule, land was common property 
of the village, but in regard to cultivated land we find the beginnings 
of sib, family, or individual ownership." 3 

Up to a point conditions of land tenure are similar in hydraulic 
tribes. Among the smaller irrigation tribes of equatorial Africa land 
can be bought and sold. This is the case among the Suk * and the 
Endo. 6 Among the En-Jemusi it was originally "marked out by the 
chief," but now, when division after the father's death excessively 
reduces an allotment, an owner can augment his holdings by pur- 

a. The Iroquois have a saying: "Land cannot be bought and sold, any more than 
water and fire can" (Lips, 1938: 516). 

CHAPTER 7, E 233 

chase, as do the Suk; or following the earlier pattern, he may be 
given additional fields by his chief. 6 In the American Pueblos com- 
munal patterns of land tenure prevailed until modern times. In the 
Rio Grande area "unused agricultural land reverts to the town, to 
be reallotted by Town chief [cacique] or Governor." 7 Among the 
hydraulically marginal Hopi a "clan system of land tenure was uni- 
versally in vogue"; 8 and the village chief, who was "the theoretical 
owner of all the village lands," * asserted his authority "most fre- 
quently ... in the settlement of land disputes." 10 

Thus in both nonhydraulic and hydraulic small farming com- 
munities the forms of land tenure vary; and the tendency toward 
communal control is strong but not universal. Corresponding resem- 
blances can be discovered also with respect to mobile property. 
Weapons as well as tools that are used in hunting and gathering 
are usually owned individually by hydraulic tribesmen; but the 
objects thus obtained are so perishable that their passing possession 
does not favor the development of class distinctions, whatever the 
methods of distribution. 

Nor, under such conditions, do industry and trade lead to sig- 
nificant social differentiations. This is eminently clear with regard 
to trade. The exchange of privately owned goods is undertaken 
privately; but this does not require special training or full-time 
handling. As in the small nonhydraulic farming communities, there 
is trade in hydraulic tribes, but there are no professional traders. 6 

2. Specific Aspects 

In industry conditions are not so simple. Property-based crafts are 
practiced primarily to satisfy the farmers' personal needs; and those 
who, because they command particular skills or have access to par- 
ticular materials, produce goods for exchange usually do so on a 
part-time basis, their major efforts still being devoted to agriculture. 
This is the prevailing pattern in both nonhydraulic and hydraulic 
tribes, and it is a pattern that is not fundamentally altered by the 
presence of a few professional craftsmen, such as smiths. 

b. Among the Pueblo Indians exchange between the various villages or with non- 
Pueblo peoples is maintained by individuals (Parsons, 1939, I: 35; Beaglehole, 1937: 
81) or by trading parties (Parsons, 1939, I: 34 ff.). Market-like gatherings are organized, 
usually by women (Beaglehole, 1937: 8a ff.; Parsons, 1939, I: 36 ff.) and, it seems, 
spontaneously (Beaglehole, 1937: 81 ff.). For earlier conditions see Espejo, 1916: 183; 
Bandelier, FR, I: 101, 163; Parsons, 1939, I: 33 ff.; Hackett, 1923, II: 234, 236, 240, 
242 ff.; for recent developments, see Parsons, 1939, I: 34 ff. For the Chagga see 
Widenmann, 1899: 69; Gutmann, 1926: 425, 431. 

c. Beech, 1911: 18. The potters mentioned by Beech (p. 17) obviously give only 
part of their time to their craft. 


Large-scale constructions are a different matter. Small farming 
communities of the nonhydraulic type usually lack the organizational 
integration for the execution of such enterprises; and some hydraulic 
tribes, such as the Suk and the Endo, have not applied the organiza- 
tional methods they employ in hydraulic work to nonhydraulic ob- 
jectives, as the American Pueblo Indians have done with amazing 
success. To be sure, the tools of the Pueblo builders were privately 
owned; but their building materials were secured under communal 
leadership, and the work was done by communal labor. Such arrange- 
ments do not promote a property-based private industry nor the 
growth of a group that derives its strength from private industrial 
property and enterprise. On the contrary. They clear the way for 
patterns of operation that retard the rise of nongovernmental pro- 
prietary forces in industry as well as in other sectors of society. 

In the sphere of hydraulic works these antiproprietary forces ap- 
pear regularly. A primitive peasant, using his own tools, cultivates 
land that may or may not be communally regulated, and the seeds 
for his crops may belong to him personally or to his kin group. Under 
nonhydraulic conditions this is the whole story. In a hydraulic setting 
cultivation proper follows a similar pattern; but the "preparatory" 
operations do not. The tools are privately owned, but the raw ma- 
terials for making the hydraulic installations (earth, stone, and per 
haps timber) either are communal property — that is, owned by nobody 
or everybody — or, if they are to be found on land held by a particular 
individual, family, or clan, are taken over by the community. And 
the end products of the community's coordinated effort, the ditches 
or canals, do not become the property of the individual farmers or 
farming families that participate in the work, but, like the water 
which they carry to the individual fields, they are controlled ("owned") 
by the community's governing agency.'* This proprietary peculiarity 
can be discerned in the incipient hydraulic communities of the Hill 
Suk, whose "irrigation ditches are the property of the tribe, not of 
the individual." u In the irrigation villages of the En-Jemusi the 
irrigation ditches are also the property of the tribe; 12 and this is 
equally the case with the larger, communally built, irrigation instal- 
lations of the Pueblo Indians. 

To evaluate these facts properly we must remember that the com- 
munities discussed so far are small farming societies — that is, com- 
munities in which the basic unit of tribal activity is almost always 
the village. In a nonhydraulic setting the headmen of the small units 
do not, as a rule, have authority over any substantial communally 

d. Small ditches that require the labor of only a few individuals or a kin group are 
the property of those who make them. 

CHAPTER 7, E 235 

owned and communally managed property. Such property, however, 
characterizes the hydraulic village; and in most cases it is administered 
by ceremonial and/or operational leaders/ 

This proprietary development has another aspect, which has al- 
ready been noted but which in the present context assumes a new 
significance. In small nonhydraulic farming societies a headman, who 
exerts little functional leadership, does not have his fields tilled for 
him by the community. Among small hydraulic tribes the headman, 
even when his leadership is overtly recognized, is not always so 
privileged either/ However, among the Pueblo Indians, who in most 
cases combined a compact hydraulic agriculture with large nonhy- 
draulic constructions, the chiefs fields were cultivated for him, even 
in villages that numbered only a few hundred inhabitants. 

Among larger hydraulic tribes, such as the Chagga, the existence 
of the chief's fields cannot be considered specific, since such land 
arrangements occur in large nonhydraulic communities. But in large 
hydraulic tribes the chief's fields tend to be extensive; and work on 
them (and on the chief's houses) is done not by a limited number of 
retainers but by all able-bodied tribesmen." Another proprietary 
peculiarity is entirely specific: the chief's privileged claim on the 
tribe's irrigation water. 13 

The extraordinary concentration of land, water, agricultural and 

e. For the Pueblos the directing authority of the cacique and the war chief is well 
established. The situation among the Hill Suk is less clear. Beech (1911: 15) recog- 
nized that communal discipline was invoked in hydraulic work, but he was unable to 
discover any directing secular leader, or for that matter any religious leaders: 
"medicine men" (ibid.: xiv, n. 1). However, an "Elder" plays a prominent role in 
two crucial agricultural ceremonies, one pertaining to the clearing of the land, the 
other to the opening of the irrigation ditches (ibid.: 15 ff.). Sir Charles Eliot doubts the 
validity of Beech's anarchistic picture (ibid.: xiv, n. 1); and he does so by citing military 
requirements. No doubt the need for military leadership exists in almost all inde- 
pendent communities, but Eliot's military argument would be equally valid for the 
small nonhydraulic farming communities, whose chiefs rarely have more than a 
"purely representative position" (Lips, 1938: 515). In the Pueblos tribal leadership 
is definitely linked to leadership in communal activities, and among them hydraulic 
work ranks first. Expanding Eliot's reservations, we suggest that the germs of an 
operational authority were present among the Hill Suk, particularly in the matter of 
the tribe's most important property, its hydraulic installations. 

/. The chief occupies a conspicuously strong position among the En-Jemusi (Beech, 
1911: 37), but there is no evidence of any public fields being tilled for him. 

g. The Chagga chieftain demands corvee labor from the tribe's adult males, from 
the women, and from the adolescent boys. These three groups work for the chief, 
in agriculture: cutting the bush (men), burning (men), hoeing (women), watering 
the seeds (men), raking and weeding (women), irrigating (men), and harvesting 
(women) (Gutmann, 1926: 376); in construction work: cutting and transporting timber 
(men), building proper (men), carrying of heavy loads of straw for the roofs (women), 
bringing up material for fences, etc. (the "boys") (ibid.: 376, 368). 


industrial labor in the hands of the chiefs does not enhance personal, 
family, or clan ownership.* It does not benefit the social position of 
private craftsmen, who in the larger hydraulic tribes become some- 
what more numerous. 4 Nor does it favor private professional mer- 
chants.' Specifically, it hampers the expansion of private property in 
what is frequently an important secondary branch of the subsistence 
economy: herding. 

The tribal history of many European civilizations shows how, in 
an agrarian economy, growing cattle wealth is a factor in establishing 
societal leadership. In East Africa animal wealth is similarly esteemed; 
and in a predominantly pastoral community, such as the Masai, this 
wealth, which is eagerly displayed, 14 is an essential means of deter- 
mining the owners' social position.* Not so among the Chagga. Cattle, 
which under the peculiar conditions of the Chagga area were largely 
stall-fed, 16 increased substantially; and some tribesmen owned as 
many as eighty head. 16 But in Chagga society the owners of large 
herds did not necessarily enjoy a higher social status, although they 
certainly enjoyed added material advantages. The Chagga chieftain, 
thanks to his quasidespotic powers, easily found a pretext for ac- 

h. Until recent colonial times the bulk of all Chagga land was controlled, first by 
the clans and subsequently, and increasingly, by the chieftain. The clans yielded to 
the chief some of their authority over the banana lands, which were probably the first 
to be cultivated and required some irrigation (Gutmann, 1926: 303; Dundas, 1924: 
300 ff.). The fields of eleusine millet, which had always required intensive irrigation 
"are marked out and allotted by the Chief himself. So are the maize fields in the 
plains, and this allotment is one of the important duties of the Chief" (ibid.: 301). For 
recent colonial developments in the chieftain-controlled maize sector see Gutmann, 
1926: 307. 

i*. Among the Chagga the only professional craftsmen are the smith and perhaps 
the tanner (Widenmann, 1899: 84; Gutmann, 1909: ng; Dundas, 1924: 270 ff.). The 
smiths live in special localities and they may only marry women from families of smiths 
(Widenmann, 1899: 84; Gutmann, 1909: ng; Dundas, 1924: 271). 

;'. Among the Chagga, even more exclusively than in the Pueblos, trade is in the 
hands of women (Widenmann, 1899: 69; Gutmann, 1926: 425). 

k. Merker, 1904: 28. Among the Pastoral Suk, who "rather look down upon them 
[the Agricultural (Hill) Suk] on account of their poverty" (Beech, 1911: 15), cattle 
wealth seems to be decisive for the establishment of communal prominence. A certain 
Karole, who had the reputation of being the "richest" of the Suk (ibid.: 7, n. 1), rose 
as high politically as the undifferentiated conditions of his tribe permitted; he became 
his group's "most important advisor" (ibid.). But the overt authority of the "advisors" 
was extremely slight; and it is doubtful whether, among the Pastoral Suk, any of 
them exerted more power covertly, since no communal enterprise known to us pro- 
vided an opportunity for invoking generally accepted disciplinary methods. It is 
probably no accident that the poorer but incipiently hydraulic Hill Suk prosecuted 
persons who violated the tribal laws more severely than did the wealthier plainsmen: 
"The punishments for crime in the hills are far stricter than in the plains" (ibid.: 27, 
n. ,). 

CHAPTER 7, E 237 

cusing conspicuous cattle owners of some malfeasance or other and 
for confiscating some or all of their animals." And the Chagga herders, 
instead of boasting of their growing cattle wealth, became increas- 
ingly secretive and fearful. An earlier practice of farming out cattle 
to poorer tribesmen for foddering 18 became a convenient device for 
hiding their valuable but insecure property. The animals were now 
handed over to their temporary keepers furtively and by night; 19 and 
the owners' sons, who originally had played an important role in 
the transfer, 20 were at times not even informed as to where the cattle 
were placed. Says Dundas: "So secret does he keep the whereabouts 
of his stock, that he will not even tell his sons where it is." 21 This 
trend gained strength with the growth of chiefly power, which oc- 
curred prior to the establishment of colonial rule. It was further 
aggravated when, under this rule, the chief started to raise a general 
cattje tax. 22 

In this setting private wealth does not necessarily, or even primarily, 
establish public prominence."* Among the qualities that in the earlier 
time favored chieftainship, wealth probably was a desirable but not a 
necessary factor; and the chief's property certainly grew not in pro- 
portion to what wealth he or his forefathers may have had originally 
but in proportion to his growing agromanagerial and military power. 
For his aides the ruler chose men who were prominent in their 
locality 23 or — and increasingly — men whose personal qualifications 
fitted them for the job. 24 In both cases selection involved a conspicu- 
ous improvement in the material conditions of the chosen, for the 
chief provided his serving men with cattle and women. 25 In fact, 
Merker found that only persons in government positions were rich. 26 

3. Simple I . . . 

Manifestly, hydraulic tribes like nonhydraulic agrarian tribes 
develop private property. Both conformations present undifferenti- 
ated forms of property (as in handicraft and trade) and a trend to- 
rn. Gutmann (1909: 7) says that rich tribesmen may withhold irrigation water from 
the poor, but in a later and more detailed study he describes the equalitarian way 
in which all members of a given hydraulic unit are provided with water (1926: 418). 
He also refers to certain "nobles" who obviously owned cattle and who helped a 
chieftain obtain office (ibid.: 462). But no details are given regarding this incident, 
which occurred at the beginning of the 19th century (ibid.: 461), that is, before the 
chief's leadership in communal affairs had been fully established. And the clan 
leaders did not owe their rank to wealth, though, once chosen, some of them probably 
had opportunities for improving their economic condition (ibid.: 15). A clansman 
became ceremonial leader because he was the oldest male in the group (ibid.: 13), and 
the political leader, the "speaker," achieved his position on the basis "not of his age, 
nor of his wealth, but of his political shrewdness" (ibid.: 14). 


ward regulated forms (as in farming with respect to land). At the same 
time, however, significant differences may be observed. Under hy- 
draulic circumstances, political property already emerges in small 
hydraulically compact communities (the chiefs land in the Pueblo 
villages). In larger tribes political property expands one-sidedly, and 
it retards and cripples private property in important spheres of ac- 
tivity (such as herding). 

The difference between this one-sided accumulation of property 
in the hands of the governing authorities and the pluralistic patterns 
of proprietary growth in nonhydraulic agrarian tribes reflects per- 
fectly the differences in the character and weight of political author- 
ity." In the German tribes observed by Caesar and Tacitus the 
chieftain, although recognized as the top-ranking political leader and 
expected to devote much of his time to his governmental duties, was 
unable to restrict or tax the wealth of his nobles. Nor did he demand 
corvee labor or taxes from his tribesmen, who would have considered 
such a request an insult and who, like the nobles, participated in the 
public discussions of the tribe's affairs. 27 

Thus in tribal hydraulic societies property is simple, but it is 
simple with a specific tendency toward the predominance of political, 
power-based, property. This tendency increases with the size of the 
community. It becomes decisive in simple hydraulic commonwealths 
that are no longer directed by a primitive (tribal) government, but 
by a state. 


1. Statehood versus Primitive Government 

Control over a distinct territory has been considered a basic aspect 
of statehood. This aspect is indeed essential; but it has little value 
in the present context, since it is not specific. (As a rule, primitive 
governments also claim control over their territory.) Nor does the 
criterion of sovereignty help much. (Primitive governments also 
strive to establish sovereignty; and like states, they are not always 
able to.) 

The differences between a primitive government and a state seem 
inconsequential so long as we confine comparison to external rela- 
tions. They become significant when we compare internal conditions. 

n. As elaborated above, in most nonhydraulic communities tribal coordination is 
required mainly for military and ceremonial purposes, whereas the heads of hydraulic 
tribes, in addition to exerting military and/or religious leadership, fulfill specific, vital 
agromanagerial functions. 

CHAPTER 7, F 239 

Primitive governments are operated in the main by nonprofessionals 
— that is, by functionaries who devote the bulk of their time not to 
the civil, military, or religious affairs of the community but to their 
own hunting, fishing, farming, or raiding. States are operated in the 
main by professionals — that is, by functionaries who devote the bulk 
of their time to "public" affairs. From the standpoint of human rela- 
tions a state means government by professionals. 

Certain communal functions, such as the maintenance of internal 
order and the organization of defense, are vital for the perpetuation 
of all types of society. Consequently, man's political activities are 
as essential as those involved in the securing of food and shelter; and 
the professionalizing of government is as important an aspect of 
social differentiation as is the professionalizing of those economic or 
intellectual pursuits that under more primitive conditions are handled 
only by persons who are primarily engaged otherwise. 

It goes without saying that a statelike government with its full- 
time civil and military officials, its soldiery and police can invest 
much more time and energy in administrative and coercive activities 
than a primitive government. It is this power potential of the state 
that makes its control by responsible and effective nongovernmental 
forces the only guarantee against the rise of a totally powerful (and 
totally corrupt) apparatus state. 

Many Marxists, following Marx' and Engels' interpretation of the 
Western state and disregarding their stress on the peculiarity of 
Oriental despotism, have described "the state" as an institution that 
always serves the special interests of a property-based ruling class. This 
interpretation, which today, in its Soviet version, is part of an ex- 
tremely widespread — and extremely potent — political myth, is not 
true even for modern parliamentary governments, whose plutocratic 
potential it generalizes and whose capacity for growth and democ- 
ratization it denies. Nor does it fit the states of Western absolutism 
and feudalism, nor indeed the democratic states of ancient Greece. 
And it is completely absurd when it is applied to the agrarian and 
industrial apparatus states that are characterized not by the strong 
influence of nongovernmental proprietary forces on the state, but 
by the abysmal lack of any such influence. 

2. Steps in the Professionalizing of Government 

a. Chagga Chieftainship and the State of Ancient Hawaii 

The difference between primitive government and statehood be- 
comes unmistakably clear when we juxtapose the single full-time and 


community-supported leader of a Pueblo village and the large staffs 
of government functionaries in Pharaonic Egypt, imperial China, or 
Ottoman Turkey. The almost complete predominance of nonpro- 
fessionals in the first case is as manifest as the almost complete pre- 
dominance of professional apparatus men in the second. The 
difference is less obvious, but perhaps even more informative, when 
we compare the regimes of large hydraulic tribes, such as the Chagga, 
with the state of a relatively crude neolithic hydraulic civilization, 
such as ancient Hawaii. 

The absolutist acts of a Chagga chief are impressive: he kills, ° 
spies, seizes his subjects' cattle, 6 and keeps as many girls as he wishes 
in his palace." In addition and more importantly, he is the com- 
mander in chief of the tribe's laboring and fighting force.* Neverthe- 
less, his ability to rule the lives of his subjects is limited by the small 
number of his full-time functionaries. Highest among them is "a 
person who may best be described as his prime minister, and on 
whom much of the executive work devolves." * Below this tribal 
version of a vizier are certain helpers and advisors, akida, 2 who "re- 
ceive the chief's orders, convey them to the people, using for this 
purpose special helpers, and supervise and organize their execution. 
Such orders pertain, for instance, to the making and repairing of 
canals, work for the chieftain . . . payment of taxes and religious 
affairs." 3 The akida, who are expected to spend a considerable part 
of their time at the chief's palace,* apparently have one assistant each; B 
but the professional officialdom ends here. Clan heads may advise 
the chief, 8 staying at his palace for this very purpose, and most of 
the on-the-spot directing remains in the hands of the clans. The 
hornblower, the actual leader of the corve>, is selected by members of 
his clan and only confirmed by the chieftain. 7 Obviously, he is 
not a full-time salaried functionary. 8 

a. To demonstrate his loyalty a Chagga dignitary was ready to burn his sister to 
death when ordered to do so by his chieftain (Gutmann, 1914: 219). 

b. As punishment for an alleged crime, Chieftain Mapfuluke is said to have seized 
the cattle of one of his fathers-in-law. Later, and quite unexpectedly, he returned some 
part of them (Gutmann, 1914: 231). 

c. Gutmann (1926: 388 ft.) estimates that, in one instance, the chieftain assembled 
from among rank-and-file families more than 5 per cent of all girls. These young 
females were then assigned to his wives; but the chieftain maintained his sexual rights 
over all of them: "None of the girls entered marriage untouched, the chieftain used 
them as he pleased." 

d. The Chagga chieftain makes the supreme decisions concerning the hydraulic 
corvee and other large-scale secular enterprises. He commands his tribesmen in war; 
he assigns residences to all; and he fixes the dates for sowing and harvesting (Gutmann, 
1909: 25). 

CHAPTER 7, F 241 

Nor does the chieftain have at his disposal professional guards or 
policemen. The warriors who protect his person — and this is par- 
ticularly demanded at night — are ordinary members of the tribe who 
return home after their shifts are done. 9 

The supreme head of the Chagga government is occasionally re- 
ferred to as a "monarch" or "king." e However, the majority of all 
observers designate him as "chieftain." 10 Conversely, the ancient 
Hawaiian rulers are sometimes called "chiefs," but in more scholarly 
treatises they are designated as "kings." The preferred titles reflect 
the general conviction that the Chagga ruler presides over a more 
primitive type of government than does his Hawaiian counterpart. 
This conviction seems well founded. In the first case we are faced 
with a primitive government that has elements of incipient state- 
hood, in the second with a crude but genuine state. 

The Hawaiian kings disposed over a much more differentiated 
staff of top-ranking aides than did the Chagga chieftains. In addition 
to a chief councilor, the Hawaiian ruler had a chief war leader, a 
chief steward, a treasurer, and "land experts." X1 There is no evidence 
that clan heads acted as his advisors or that his guards served part 
time. Besides a "body guard," the king had at his beck and call a 
detachment of armed men headed by an executioner — official ter- 
rorizes who were always ready in the king's name to accuse, arrest, 
and kill. 12 

In the Hawaiian government the professionals were not confined 
to the top echelon. Below the leading officials there were primarily 
and most importantly the konohiki. In contrast to the Chagga akida, 
who spent much of their time near their chieftain, the konohiki 
seem to have resided and officiated for the most part in the regions 
of their jurisdiction, directing the regime's constructional, organiza- 
tional, and acquisitive operations. They kept count of the popula- 
tion; 13 they mobilized the corvee; " they directed the hydraulic enter- 
prises; 15 they supervised agriculture; 16 they gathered the tax, 17 
retaining some of it for their own use and for the use of their under- 
lings, but passing on most of it to the higher authorities, and eventu- 
ally to the king. 18 

Manifestly, the konohiki and their aides were government-sup- 
ported full-time functionaries. The organizational and acquisitive 
network that they spread over the countryside probably contributed 
more than any other political institution to making the government 
of ancient Hawaii a crude, agrobureaucratic hydraulic state. 

e. Gutmann, 1909: 10 ff. Lowie (1938: 30s) calls him the head of a "monarchical 



b. Proprietary Consequences 

Controlling a much more fertile territory and a much larger 
population — the largest Hawaiian kingdom had five times the popu- 
lation of the largest Chagga tribe ' — the Hawaiian rulers were in 
a better position to establish and maintain a permanent officialdom. 
And this larger officialdom in turn enabled them to control their 
subjects' property more completely. In Hawaii the government's 
jurisdiction over the land was not restricted by any clan rights, as 
was the case among the Chagga. 19 Nor did a clan head stand between 
tax-collecting officials and individual taxpayers, as in Chaggaland. 20 
Indeed the Hawaiian regime functioned so well that the masters of 
the apparatus state were able to syphon off over half of the entire 
rural produce. According to one estimate, "the common laborers did 
not receive on an average more than one third of the avails of their 
industry." g 

On a smaller scale the difference between the two types of govern- 
ment appear also in the sphere of circulation. The Chagga markets 
were policed by the chief's wives and regional officials; 21 but a mar- 
ket tax on agricultural products and salt was collected by a member 
of one particular clan." In Hawaii we find no trace of such divided 
authority. The functionaries who sanctioned the transactions and 
taxed the goods were toll collectors — that is, government officials. 22 

Thus the kings of Hawaii exerted a much more formidable power 

/. In the 18th century some 300,000 Hawaiians were organized in a few sovereignties, 
the largest of which, Hawaii proper, numbered more than 85,000 people (Lind, 1938: 
60). Lind's figure harmonizes well with an estimate made by Ellis in 1826 (Ellis, 
1826: 8). Ellis considered the total of 400,000 inhabitants suggested by the earliest 
observers "somewhat above the actual population of that time, though traces of 
deserted villages, and numerous enclosures formerly cultivated, but now lying waste, 
are everywhere to be met with." In 1826 there were between 130,000 to 150,000 people 
on the archipelago (Ellis, 1826: 8). Fornander, although suggesting smaller figures than 
Cook and King, sees "no valid reasons for assuming a greater or more rapid depopula- 
tion between 1778 and 1832, when the first regular census taken gave an approximately 
correct enumeration of 130,000 than between the latter year and 1878, when the 
census gave only 44,088, exclusive of foreigners" (Fornander, PR, II: 165). At the 
beginning of the 19th century Bali had a population of about 760,000 persons, with 
some of the island's major kingdoms accounting for more than 100,000 persons each 
(Lauts, 1848: 104-5). Tlie largest Chagga tribes numbered less than 20,000, 10,000, or 
5,000 persons respectively (Gutmann, 1926: 1). 

g. Alexander, 1899: 28 n. Blackman (1899: 26) presents this estimate as "the 
opinion of careful observers." 

h, Gutmann, 1926: 426 ff. The clan functionary grasps a handful of taxable goods. 
The trading women have the right to kick him once; but they cannot prevent him 
from seizing the fee, which "at a well attended market amounts to fairly sizable 
loads" (ibid.: 427). 

CHAPTER 7, F 843 

over their subjects' life and property than did the Chagga chieftains. 
Difference in the form of reverence strikingly expresses the difference 
in autocratic power. As already mentioned, the Chagga tribesmen 
hold their ruler in great esteem, but unlike the Hawaiians they do 
not perform before him the classical gesture of total submission: 

3. Simple Patterns of Property in Land, Industry, 
and Commerce 

In the early phases of state-centered hydraulic societies, private prop- 
erty in land is not necessarily lacking; its origins go back much 
further than was assumed by the pioneer institutionalists of the 19th 
century. But the greater part of all cultivable land is regulated, and 
thus kept from being privately owned, even after private and inde- 
pendent property emerged notably in industry and commerce. For 
this reason, we shall discuss the problems of hydraulic land tenure 
later. With respect to the simple patterns of hydraulic property we 
need only state here that, within the framework of these patterns, 
the forms of land tenure are many but that regulated land always 
prevails (and generally by a substantial margin) over privately owned 
("free") land. 

Property-based and independent handicraft and commerce, how- 
ever, must be examined immediately, for their occurrence makes, 
as we see it, a change in the patterns of property and society. This 
development is not uniform. 
It advances unevenly in the spheres of 

A. Industry, in 

1. The extractive industries (mining, quarrying, certain 

forms of salt production) 

2. The processing industries 

a. Constructions 

b. Others 
and also in 

B. Commerce, in 

1. Foreign trade 

2. Domestic trade, dealing with 

a. Easily supervised goods (such as salt, iron, tea, wine, 
oil, etc.) 

b. Others. 

In all hydraulic societies proper and in most marginal hydraulic 
societies, the government engaged in comprehensive constructions. 
Employing a large labor force, the agrarian apparatus state enjoys 
what amounts to a monopoly of all large-scale construction work. 


Often it also manages those extractive operations which provide the 
bulk of all raw materials for the large government constructions. 
Other extractive industries, such as mining and certain forms of 
salt production, may either be directly managed by the government 
or, and particularly under the conditions of a money economy, they 
may be controlled through monopolistic licensing. 

Thus property-based and independent action cannot hope to pre- 
vail in the most important sector of hydraulic industry: large-scale 
constructions. Nor can it hope to operate freely in the large extractive 
enterprises. Only in the nonconstructional sector of the processing 
industries is there a chance for property-based free handicraft to 
become significant. Indeed, apart from the making of coins, only 
a few manufacturing pursuits, such as the production of weapons 
and certain luxury goods, may be directly managed by the govern- 
ment, while most other crafts are handled entirely by private and 
independent entrepreneurs. 

Free private enterprise, however, does not necessarily mean large 
enterprise. Large-scale industries are extremely vulnerable on the 
fiscal level, and except for government-protected units do not prosper 
under the shadow of total power. The many private and independent 
crafts which have emerged in certain hydraulic societies are essentially 
confined to small shops and small-scale operations. 

The development of private big trade may be retarded under 
conditions of great hydraulic and bureaucratic density (compact- 
ness), but it is not blocked by the state's managerial predomi- 
nance, which, with regard to the construction industries, appears in 
all hydraulic societies proper and also in many marginal hydraulic 
societies. Above the level of the producer-trader, commercial business 
is transacted over significant distances, either overland or oversea. 
This favors large-scale action, particularly since the merchandise so 
handled is less conspicuous and therefore less vulnerable fiscally than 
a fixed and conspicuous industrial plant. 

When the law of diminishing administrative returns induces a 
state to limit its own commercial operations, independent merchants 
tend to appear both in foreign and domestic trade; and governmental 
attempts to maintain direct and indirect controls in both sectors at 
a particular level or to restore them to an earlier level are based for 
the most part on short-range considerations.* 

i. This is why government policy in this regard in China, India, and the Near East 
oscillated so considerably. The student of Chinese history will recall the discussions 
which Han administrators conducted concerning the way in which the sale of salt 
and iron should be handled. The problem arose in pre-Han days, and different 
solutions were found at different times. The administrative history of India is not 

CHAPTER 7, F 245 

Hydraulic society outgrows the simple patterns of property, when 
private and independent handicraft becomes prominent in the pro- 
cessing industries (excluding, of course, large-scale construction) and 
when big and independent merchants handle as much or more busi- 
ness than all government-managed and government-controlled com- 
merce taken together. 

The almost complete absence of pertinent statistical data compels 
us to formulate our criteria broadly. In some branches the relative 
proportions are evident. In others we can at least establish prevailing 

4. Variants of Simple Patterns of Hydraulic 
Property and Society 

a. Hawaii 

The Hawaiian archipelago is so distant from the more southerly 
regions of the Polynesian world that after an early period of daring 
expeditions, "all intercourse with the southern groups seems to have 
ceased, for there is no further evidence of it in any of the ancient 
legends, songs, or genealogies for five hundred years." 23 

Nor were the relations between the various Hawaiian kingdoms 
sufficient to stimulate the development of commerce above the pro- 
ducer-trader level. 21 Internal circulation consisted in the main in 
the transfer of rural surpluses from the peasant and fishermen pro- 
ducers to the local and central representatives of the government. 
Exchange between individuals occurred either in the form of 
"gifts" 25 or barter; 26 and in both instances without the aid of pro- 
fessional middlemen. Markets and fairs provided ample opportunity 
for such activity. Ellis' descriptions of what was then considered 
the most famous fair makes no reference at all to any professional 
merchants. The only professional person noted by the observer 
was the government official who supervised and taxed the transac- 
tions between the barterers. 27 When, in the early 19th century, con- 
tact with the outside world opened up a new outlet for sandalwood, 
it was the king and his lieutenants and not independent private 

so well documented as that of China, but what we know about Indian fiscal policy 
suggests similar oscillations. 

The history of state and private commerce in the great hydraulic countries of the 
Near East is still in its infancy; and attempts, such as that made recently by Leemans, 
reveal the institutional importance of this phenomenon as well as the difficulties of 
investigating it. The Near Eastern data show again that in contrast to the great 
hydraulic works and the big nonhydraulic constructions, large-scale commerce can 
readily be handled by private and independent merchants. 



Hawaiian merchants who handled the resulting international trade. 

The undeveloped conditions of circulation reflect the undeveloped 
industrial conditions, and these in turn are closely connected with 
the paucity of suitable raw materials. The volcanic islands of Hawaii 
lack metals; and this deficiency kept the islanders, as long as they 
were separated from technically more advanced civilizations, at a 
relatively crude level of neolithic life. The archipelago had useful 
plants (such as taro and the coconut tree) but none of the world's 
major cereals; and there were no animals that could be used to ease 
man's labors. Lava was the only important workable stone. 

The technical skill which the Hawaiians developed in this natural 
and cultural setting was admirable. 29 However, even maximum in- 
genuity produced only a modest differentiation in the crafts. Special- 
ists built canoes 30 and houses, 31 made nets, fish lines, tapa cloth, 32 
and many other articles, 33 yet the economic and political position 
of these artisans is none too clear. A number of them may well have 
worked for their own account.' But neither Hawaiian tradition nor 
early non-Hawaiian observers suggest that these private artisans 
could compare in importance with the craftsmen who served the 
king and his functionaries. The government, which controlled an 
enormous percentage of the country's surplus, was able to support 
many artisans, poe lawelawe. The supreme poe lawelatoe was a mem- 
ber of the central government. 34 He seems to have directed the in- 
dustrial activities undertaken for the benefit of the government and 
obviously through the use of corvee labor. In addition, he was in 
charge of the numerous artisans who were permanently attached to 
the court. Says Kepelino: "At the chief's [king's] place there were 
many workers or Poe- lawelawe of every description." 35 

Thus in ancient Hawaii professional artisans appeared most signifi- 
cantly as persons who, supported by the government, worked under 
government functionaries for the ruler and his serving men. This 
constellation, together with the complete absence of independent 
professional merchants, created in ancient Hawaii a very rudimen- 
tary variant of simple patterns of hydraulic property and society. 

b. Inca Peru 

The masters of the Inca empire drew upon natural resources that 
were richer than those of Hawaii but poorer than those of Egypt, 
Mesopotamia, China, or India. The agriculturists of the Andean area 
entered the metal age at a relatively late date; and even then they 
did not process iron. Nor did they domesticate animals for use in 
farming. To be sure, in hydraulic civilizations the absence of labor 

;'. Several trades had special patron gods (Alexander, 1899: 37, 62 ff.; Blackman, 
1899: 38). 




















CHAPTER 7, F 247 

animals is of less importance to crop-raising k than to transporta- 
tion, which is basic to the spread of military and political control, 
to the collection of taxes, and to the growth of trade. However, 
compared with the donkey, the mule, the ox, the horse, and the 
camel — the chief labor animals of the Old World — the llama, al- 
though useful for its wool, was a poor instrument of locomotion. 
The absence of navigable rivers, in addition to a rugged coastline, 
discouraged experiments in shipping, except on primitive rafts; and 
a scarcity of culturally advanced neighbors discouraged international 
trade much more decisively than was the case in Pharaonic Egypt. 

Table 5. Factors Stimulating Commerce and Regional Division of 

Labor in Industry 


Inca Peru 

Pharaonic Egypt (particularly the 

Old and Middle Kingdoms) 
The various states of ancient China 

Key 1. The ox was used for plowing only 

-f Present at the close of the Chou period. 

— Absent 
( ) Development limited 

Our analysis has revealed a number of factors that stimulate com- 
merce and regional division of labor in industry. We indicate in 
Table 5 the uneven development of these factors for a number of 
major simple hydraulic civilizations. Although by no means the only 
formative features, they aid us in recognizing the uneven develop- 
ment of trade and industry in these civilizations. 

In the Andean area transportation was further discouraged by 
the desert-like conditions in large segments of the coast and by the 
high and steep elevations in the strategically located mountain re- 
gions. For all these reasons, effective and long-distance communica- 
tion proceeded essentially on land and not on water; and it depended 

k. An approach which recognizes the crucial role of hydraulic operations in the 
development of agriculture cannot be content with Lowie's otherwise suggestive 
typology of subsistence economies: "hunting, farming with hoe or dibble, farming 
with plow and livestock, and stockbreeding without farming (pastoral nomadism)" 
(Lowie, 1938: 283). The Near East, India, and China shared the plow and labor 
animals with Europe and Japan; and the reason for the differences between the 
stationary hydraulic civilizations and other agrarian civilizations that were not 
stationary must therefore be looked for elsewhere, and most decisively, it would seem, 
in the presence or absence of hydraulic agriculture. 


to an extraordinary degree on roads that were built and controlled 
by the omnipotent hydraulic state. There were a few foreign-mer- 
chants; 36 and some of the trade in salt and fish reported for the 
northern border zone 37 may have been handled by professionals. 
But such developments were so peripheral and of so little importance 
that serious scholars, such as Means, have completely overlooked 
their occurrence. Within the empire government officials directed 
the transfer of enormous quantities of goods — corn, beans, cotton, 
timber, metal, textiles, etc. — along the coast, on the altiplano, and 
from one zone to the other; and small producer-traders exchanged 
products by barter at the many fairs that were held regularly through- 
out the country. 38 But there is no evidence that any private agency 
competed with the government in long-distance transportation and 
distribution of goods. Trade there was, and on a local level obviously 
plenty of it. But there were almost no independent professional 

The industrial sphere of Inca life was much more differentiated, 
but the private artisans remained inconspicuous in comparison 
with government-employed craftsmen. The mines were managed 
either by the local heads of formerly independent territories or by 
nonlocal members of the imperial officialdom." 1 In both cases they 
were controlled by professional officeholders who, in one way or 
another, were part of the over-all agromanagerial apparatus. 

More precise information exists concerning certain aspects of the 
processing industries. The large construction teams were directed 
by prominent Inca functionaries; and the work patterns of Hawaii, 
Pharaonic Egypt, and early China suggest that here, too, special 
officials may have been in charge of the permanent government 
workers and those craftsmen who, for two months or "at most" three 
months," rendered industrial labor service in the state workshops. 
Among the permanent craftsmen that the government occupied 
there were apparently many silversmiths 39 and also not a few car- 
penters. 40 Weavers, shoemakers, lumbermen, and makers of copper 
tools are mentioned as working at home after having fulfilled their 
corvee obligations. 41 Garcilaso's description does not make it clear 
whether all, or most, of these last worked exclusively at their speciali- 
ties or whether some — or even most of them — were farmer-artisans. 
If we assume that most of them were professional craftsmen, it is 
even more noteworthy that the early accounts of rural and urban 

m. Local mining of gold in accordance with directions from Cuzco is indicated by 
Polo de Ondegardo (1872: 70 ff.). Cf. Cieza, 1945: 269; Sarmiento, 1906: 100; Rowe, 
1946: 246; Garcilaso, 1945, 1: 253; Sancho de la Hos, 1938: 181. 

n. Notable overtime was deducted from the following year's corvfe labor (Garcilaso, 
1945, I: 255). 

CHAPTER 7, F 249 

life do not mention them. It was only as permanent workers for 
the state or as members of the industrial corvee that the artisans 
became a conspicuous feature of Inca society. 

The "virgins," who were selected by officials from among the 
young and attractive females of the empire, provided the regime 
with a unique, but eminently useful, labor force. The "Selected 
Ones" were kept under strict supervision in special houses, where 
they spent the greater part of their time weaving, spinning, 42 and 
preparing beverages. The sovereign included some of them in his 
harem; and he assigned others to prominent dignitaries. But there 
were always large numbers of them confined to the "houses." Ap- 
parently there were many such establishments in the Inca empire: 
some had two hundred inmates, 43 the one in Caxa had five hundred, 4 * 
the one on Lake Titicaca one thousand, 45 and the one in Cuzco 
usually more than fifteen hundred. 46 Economically, the Inca "houses" 
constitute an interesting parallel to the textile shops of 17th- and 
18th-century Europe. Few of these latter employed more persons, 
and those employed were in the main women, who often worked 
only part of the year. 47 

Despite a not inconsiderable technical development, Inca society 
developed no conspicuous, independent, private-property-based 
classes. The sinecure land, which the Incas assigned to certain mem- 
bers of the ruling group, created no full-fledged landownership; 48 
and professional private enterprises were virtually absent in the 
spheres of transport and trade, which in other civilizations favored 
the rise of independent rich merchants. Professional private artisans, 
who certainly existed, remained an insignificant force even in the 
processing industries, when compared with the numerous craftsmen 
who permanently or temporarily practiced their skills in the govern- 
ment workshops and "houses." An interesting if feeble trend toward 
private handicraft notwithstanding, the Inca empire represents a 
simple pattern of hydraulic property and society. 

o. CPLNC: 309. The two Spaniards who gave Sancho de la Hos (1938: 181) a first- 
hand report on the Lake Titicaca temple mentioned only the preparation of sacred 
wine by the women, if the chronicler recorded their story correctly. But whatever the 
accuracy of the initial report, it seems most unlikely that the thousand "selected" 
women of the Lake Temple made nothing but chicha the year round, and this in the 
classical region of llama-breeding and wool production. Our doubts are strengthened 
by The Anonimo's comment on the dual activities of the women at Caxa (CPLNC: 
309) and by Garcilaso's description of the institution in the Inca capital. Obviously 
the virgins also had to prepare chicha and certain ceremonial foodstuffs, but their 
main operation (il principal exercicio) was spinning and weaving (Garcilaso, 1945, 
I: 188 ff.). There were many other houses of the same kind throughout the country. 
Their inmates engaged in the same' economic activities. They "spun and wove and 
made an enormous amount of cloth for the Inca" (ibid.: 189). 


c. Pharaonic Egypt 

A uniquely serviceable river provided the masters of Pharaonic 
Egypt with excellent facilities for internal communication; shipping 
was therefore well advanced at the dawn of written history. But 
scarcity of raw materials did not necessitate a regular foreign trade; 
nor was such trade stimulated by culturally advanced neighbors. The 
Egyptian ships and beasts of burden permitted the establishment of 
some external contacts, but these contacts remained intermittent — 
and essentially government managed — until the close of the Middle 

During the New Kingdom, and particularly in the days of the 
empire, private merchants emerged. But often they Were attached 
to the temples 49 and apparently they were no match for the state. 
According to Kees, during a great part of the New Kingdom the 
Pharaoh remained "the only big merchant." 50 

To be sure, foreign merchants did business in Egypt, but native 
middlemen were given even less opportunity in domestic than in 
foreign trade. 51 In the local markets producer-traders exchanged their 
goods directly, and in the main by barter. 52 A market official of the 
New Kingdom significantly bore the title "Scribe of Barter." 53 

Handicraft offered more room for the development of private 
enterprise. No matter to what extent the census data of the Old 
Kingdom imply the presence of independent trades during that 
period, p the cases of Hawaii and Inca Peru show that professional 
artisans operated in state-centered hydraulic societies that were tech- 
nically less advanced than the Old Kingdom. And a number of 
records from the Middle and the New Kingdom definitely speak of 
private artisans. 8 * 

These Egyptian private artisans were more conspicuous than their 
colleagues in the Inca empire; but, like them, they probably catered 
essentially to the every-day needs of small consumers. 55 Did they, at 
least numerically, equal the many craftsmen who in the processing 
industries were permanently or temporarily employed by the govern- 
ment and the temples? Even this is not certain. But there can be 
little doubt that economically they were less significant. 

The government engaged particularly in three kinds of industrial 
work: (1) extractive and preparatory operations requiring much 
labor, some of it skilled, but most of it unskilled; (2) big construction 
enterprises, requiring a combination of skilled and unskilled labor; 
and (3) processing industries, carried out in the main by skilled 

p. Kees (1933: 164 ff.) hesitates to accept E. Meyer's interpretation of these data 
as proving the existence of free artisans and merchants. 

CHAPTER 7, F 251 

craftsmen who were gathered together in larger or smaller workshops. 

In all three sectors the skilled craftsmen, who included artists of 
great ability, 56 seem to have been largely government employees. The 
"chiefs of work" 57 probably had supreme jurisdiction over them. In 
the branch industries they operated under specially designated fore- 
men. 58 

On the basis of carefully weighed evidence Kees concludes that "the 
economic life of [Pharaonic] Egypt constituted a not very appropriate 
soil for an estate of independent free artisans." 59 He finds the con- 
cept of free handicraft, except for lowly producers who satisfied 
lowly needs, "poorly suited for the economic picture of the Old 
Kingdom." 60 After the interlude of the Middle Kingdom, during 
which territorial courts became outstanding centers of the arts and 
crafts, 61 the New Kingdom increasingly forced the artisans into state- 
regulated workshops and subjected them to the rigid control of the 
state storehouses that allocated the raw materials. 62 

Documents from the New Kingdom show the state artisans eager 
for promotion to higher posts. Their foremen considered themselves 
fairly distinguished members of the bureaucratic hierarchy. 63 

To summarize: the power of the Pharaohs was so all embracing 
that private and independent handicraft made little headway, and 
independent professional commerce during the greater part of the 
period even less. The prevalence of state trade and the weight of 
government-managed industry, together with the dominance of state- 
regulated landed property created — and maintained — in Pharaonic 
Egypt a historically and institutionally significant variant of the 
simple pattern of hydraulic property and society. 

d. Ancient China 

The most archaic Chinese inscriptions, the divination texts of the 
Shang dynasty, mention sets of shells, which in all probability were 
used as means of exchange. But they do not clearly refer to profes- 
sional merchants. Neither do merchants play a conspicuous role in 
the inscriptions and literary texts of the Chou dynasty. Although in 
early China there certainly was trade, there seem to have been few, 
if any, professional traders. 

Big merchants, who traveled overland, are reported for the first 
part of the later Chou period, the time of the "Spring and Autumn 
Annals" (721-481 B.C.). But those on whom the data are fullest 
cooperated so closely with their rulers that they can probably be 
considered to have been government attached. 64 

During the last phase of the Chou dynasty, the time of the Warring 


States, independent merchants increased in importance — so much 
so, in fact, that in the 4th century B.C. the state of Ch'in took measures 
to restrict them. 65 By the time Ch'in had welded "all-under-heaven" 
into an empire, the great Unifier, Ch'in Shih Huang-ti, decimated the 
ranks of the merchants by sentencing them to guard the frontier, at 
first the merchants themselves and then their sons and grandsons. 66 
This policy demonstrates both the economic importance and the 
political weakness of nongovernmental professional traders at the 
end of the Chou period. 

The early Chinese records that have so little to say about profes- 
sional traders are more articulate about craftsmen. The beautiful 
bronze artifacts of Shang and early Chou reveal extraordinary in- 
dustrial refinement. However, and different from conditions in feudal 
Europe, the Chinese crafts developed not on many and separated 
manorial estates or in guild-controlled burgher towns but rather in 
big administrative centers controlled by the Son of Heaven, the ter- 
ritorial rulers, or their high ranking officials. Artisan-officials, the 
"hundred artisans," are mentioned in the oldest literary texts as 
well as in the early bronze inscriptions. 67 Apparently government 
artisans employed their skills under the supreme direction of the 
Minister of Works, the ssii-kung, && and alongside the "people," who 
as part of their corvee duty constituted the unskilled labor force of 
the government's large constructional enterprises. 

Government-attached artisans may have prevailed until the time 
of the "Spring and Autumn Annals"; 69 and perhaps it was only 
during the subsequent period of the Warring States that private 
artisans became increasingly important. 

We have no conclusive evidence that, under the Chou dynasty 
and under the first imperial dynasties, private merchants or artisans 
organized independent professional corporations (guilds) . q The re- 
tarded development in this regard is surprising when we remember 
that private handicraft and particularly private trade flourished at 
the close of and after the Chou period. Whatever the reasons for this 
unevenness, we are probably safe in suggesting that a simple Oriental 
society prevailed in ancient China until the end of Early Chou 
(722 b.c.) and probably also in the first centuries of Later Chou. 

q. Shops dealing with the same goods were apparently assembled in the same locality 
from the close of the Chou period on or from Early Han days (Kato, 1936: 79), and 
probably also prior to this time. But it "was not until after the Sui period that the 
expression 'hang', used in the sense of a street of shops of the same trade, came into 
general use"; and it was only "at the close of the T'ang period, or even later, that they 
[the Chinese merchants] came to organise a real merchants' association" (ibid.: 83). 

CHAPTER 7, F 253 

e. Sumer 

The agricultural civilizations of Lower Mesopotamia originated in 
a setting that was as lacking in certain industrial materials as it 
was encouraging to interarea exchange. The alluvial landscape, which 
because of its well-watered rivers offered ideal opportunities for hy- 
draulic development, lacked stone, timber, and metals. However, 
these materials, which were essential to technical, military, and po- 
litical growth, were available in adjacent lands, and from the stand- 
point of wealth, security, and power the incentives to obtain them 
were enormous. 

The ancient Hawaiians did not get from abroad the raw materials 
that they lacked at home; and the Andean Indians and early Egyptians 
created urban civilizations mainly on the basis of their own resources. 
The Sumerians developed a flourishing urban life, because they 
succeeded in establishing and maintaining an elaborate system of 
international relations and exchange. 

Needed raw materials can be obtained by organized force: war. 
But this is not always appropriate, and particularly not when the 
sources of supply are remote and those in control strong. In many 
cases the sough t-f or goods had to be acquired by peaceful means — 
that is, primarily by trade. 

Long distance trade requires the services of specialists in trans- 
portation and exchange. In Lower Mesopotamia merchants appeared 
early. While traders played an insignificant role in almost all other 
simple Oriental civilizations, they were conspicuously mentioned in 
the Sumerian protohistorical inscriptions of Fara; 70 and in later and 
more detailed inscriptions they were depicted as important profes- 

The development of urban centers of administration and religion 
also involved a fairly advanced division of industrial labor; and the 
Sumerian inscriptions contain many references to artisans, who prac- 
ticed their skills professionally. How developed were private property 
and private enterprise in early Lower Mesopotamia? 

Deimel's elaborate investigations suggest that from the dawn of 
history on/ the Sumerian temple cities probably offered less oppor- 
tunity for independent craftsmen than did ancient Hawaii, Peru, 
and Pharaonic Egypt. Like the other members of the temple com- 

r. According to Deimel, the ancient Sumerians apparently depended as much on 
the temples, when the Fara texts were written, as they did three or four hundred 
years later, when Urukagina ruled Lagash. "The population then too served the temple 
and lived on it" (Deimel, 1924a: 42). 


munity, the artisans received land, 71 and like them also, they rendered 
corvee service, 72 which, according to Schneider's tentative estimate, 
may have lasted some four months a year. 73 A number of craftsmen 
were employed permanently in temple workshops, 74 as were certain 
slaves (in the main female). 75 The majority of all artisans, however, 
seems to have worked for the temples through the operation of a 
putting-out system: temple storehouses provided them with raw ma- 
terials, which they processed at home and for a wage. 76 The position 
of these artisans was not unlike that of many European craftsmen who 
during the first centuries of industrial capitalism worked in a similarly 
decentralized way for their commercial or industrial employers. 

Were all domestic artisans of early Mesopotamia so engaged? And 
did any of them engage at least in some independent business? The 
second question is more easily answered than the first. The fact that 
all (or some?) of the workmen offered the temples certain tax-like 
"gifts" 77 is best explained by the assumption that they were able 
to produce something for their own account. 8 

The private activities of the Sumerian merchants were apparently 
much more extensive. No doubt these merchants were not inde- 
pendent of the city or temples either. They, too, were assigned land, 
but much more than the artisans — in fact, as much as a middle official 
or officer.' They could have their fields cultivated for them by tenants, 
wage laborers, or slaves; and their landed possessions, instead of 
handicapping them in their commercial activities, probably provided 
them with additional means for their business enterprises. As mer- 
chants, they were attached either to the supreme authority of the 
city state, 78 or to a temple, the second most important unit of power. 79 
And obviously and in the main, they traded for the "palace" or the 
temples. 80 

In their transactions the great merchants, gal damkar, and the 
ordinary merchants, damkar, enjoyed considerable freedom; 81 and 

5. A. Schneider assumes that the artisans who worked at home for the temples 
"apart from this, and perhaps already against a remuneration, also executed orders 
from other members of the temple community" (1920: 85). 

/. According to inscriptions collected by Hussey, a damkar of the Bau Temple 
received 19 gan of land (Schneider, 1920: 66). One gan could support more than one 
person and two gan a small family (ibid.: 35 ff.). A top-ranking temple executive, 
mentioned in Hussey's material, received 43 gan (ibid.: 35). Another text gives much 
higher totals for the land assigned to high officials: 90 gan and even 138% gan (ibid.). 
Heads of military detachments or other prominent warriors received 23, 24, 26, and 
18 gan, and a temple official engar 17% gan (ibid.: noff.). Among the artisans, a 
carpenter was given 1 gan, a chariot maker 1 to 2 gan, a tanner 3 gan, and cooks 
and bakers from 214 to 6 gan (ibid.). 

CHAPTER 7, F 255 

in addition they were permitted to trade for their own account. They 
might have business dealings with the ruler, 82 with the queen, 83 with 
members of the ruling family," and with less highly situated persons. 8 * 
Manifestly the opportunities for amassing wealth were vast, 85 

Thus in contrast to ancient Hawaii, China, and Pharaonic Egypt, 
Sumer saw a very early development of private enterprise in trade. 
And whereas the country's artisans, even when they were engaged 
in domestic industry, were closely tied to the temple economy, the 
merchants, who were neither trading officials nor governmental com- 
mercial agents but something in between, were much less so. Few 
simple hydraulic societies moved as conspicuously toward a property- 
based and independent commerce as did ancient Sumer. 

5. Origins of Bureaucratic Capitalism 

The great merchants of Sumer, who had funds of their own and who 
traded directly with their sovereign, occupied a position very dif- 
ferent from that of the commercial specialists of the Pharaohs. The 
representatives of the Pharaohs, who traded with Punt, 88 Phoenicia, 87 
Mesopotamia, 88 and Cyprus, 89 handled government property for the 
advantage of the government. They accomplished an exchange of 
goods often under the guise of diplomatic "presents," but they had 
a keen eye for the values involved. They asked for specific items, 90 
they carefully examined the objects offered them, 91 they criticized 
inadequate gifts, 92 and they stressed the need for reciprocity. 93 What- 
ever presents were given them during, or at the end of, their expedi- 
tions were given them as servants of the king and not as independent 
businessmen. In short, they were governmental trading officials not 
too different in their position from the members of a Soviet Trade 

In contrast to such trading officials, the government-attached mer- 
chants used their own capital largely, or exclusively, in the service 
of their rulers, who — while providing them with excellent oppor- 
tunities for doing business — might also set the conditions (prices, 
profits) under which these opportunities could be utilized. To in- 
voke a designation that originally pleased the Chinese Communists 
but that now embarrasses them, these merchants were "bureaucratic 
capitalists." 94 

In a wider sense, the designation "bureaucratic capitalists" is ap- 
plicable to several groups: (1) tax collectors, who act as fiscal agents 

u. Scholtz, 1934: 59. Princes or princesses occupied a number of artisans, servants, 
and slaves (Deimel, 1929: 126, 128; ibid., 1931: no). 


for a ruling bureaucracy; (2) officiating or nonofficiating members 
of such a bureaucracy, who on the strength of their political position 
engage in private enterprises, such as trading, money lending, and 
tax farming; (3) private businessmen, who as commercial agents or 
contractors do business for the ruling bureaucracy; and (4) private 
businessmen, who attach themselves to individual members of the 
bureaucracy to assure the success of their transactions. Bureaucratic 
capitalists, then, are owners of capital who act as commercial or fiscal 
agents for an apparatus state, no matter whether they are members 
of the officialdom or functionaries of the dominant religion, or per- 
sons of wealth who are neither. 

The records of ancient China are not clear on the subject of trading 
officials, although it seems likely that in the Shang and Early Chou 
periods certain functionaries of the early territorial states fulfilled 
commercial tasks. They are more articulate on the presence of govern- 
ment-attached commercial agents. Indeed such persons are sufficiently 
conspicuous to justify our tentatively classing Chou China, up to 
the period of the "Spring and Autumn Annals," as a simple Oriental 

For Inca Peru the very problem does not arise seriously. Officials 
of the frontier districts may have traded government-owned goods 
against goods produced abroad; and some transactions may well 
have been concluded privately. But Inca society seems to have had 
little need for trading officials and still less for government-attached 
commercial agents. 

The Sumerian inscriptions contain many references to foreign trade 
(internal exchange was mainly confined to barter). 95 Unfortunately, 
however, the texts leave many questions unanswered. What kinds of 
commercial transactions were involved in the many government ex- 
peditions that were undertaken to acquire stone, 96 wood, 97 metal, 98 
bitumen, 99 and other items? Were the majority of all merchants 
primarily trading officials or governmental commercial agents? No 
matter what the answer to these questions may be, the character of 
ancient Sumerian society provides scant justification for interpreting 
the "merchants" of the oldest inscriptions thus far deciphered as in- 
dependent entrepreneurs. 

6. The Hydraulic Sponge 

Most of the hydraulic civilizations that achieved considerable pro- 
prietary differentiation seem to have maintained simple patterns of 
property at an earlier time. In some cases, such as India, simple 
conditions of property and society gave way relatively quickly to 

CHAPTER 7, F 257 

semicomplex configurations. In other cases, such as Egypt and Lower 
Mesopotamia, they prevailed for millennia. In the Andean area they 
were (still or again?) dominant when the conquistadores arrived. 

The variations in the persistence of simple patterns of property 
assume a new meaning as soon as they are correlated with variations 
in hydraulic density. The hydraulic centers of Peru, Egypt, and 
Lower Mesopotamia all gave birth to compact systems of hydraulic 
agriculture, whereas many of the territorial states of India and China 
and, for that matter, of Mexico relied on loose or marginal types 
of Oriental agriculture. We do not, in this context, cite Hawaii, be- 
cause in that archipelago the perpetuation of extremely simple pat- 
terns of Oriental property was obviously due to an extraordinary 
combination of internal and external circumstances. However, in 
the first instances the contrast in hydraulic density patterns is too 
striking to be dismissed as irrelevant. In all probability the early 
independent hydraulic communities of the Andean zone traded be- 
yond their borders, and this early trade may well have been handled 
not only by commercial officials but also by government-attached 
private merchants, who may to some extent have acted for their own 
account. But Sumerian history demonstrates that strong hydraulic 
regimes can keep the bulk of all traders attached to the government 
even in separated city states. Thus it is not impossible that in the 
Andean area (as in Sumer and Pharaonic Egypt, but perhaps with 
more marked oscillations) there prevailed, even prior to the Incas, 
simple conditions of power, property, and class. 

In Peru these conditions may have endured as long as state-centered 
and hydraulic civilizations were present in the area. In Egypt they 
outlasted the relative isolation of the hydraulically compact Nile 
Valley. And in Lower Mesopotamia they persisted even after the 
compact hydraulic heartland had been incorporated into larger and 
looser hydraulic conformations. Leemans assumes a high development 
of private property and trade 10 ° when the second Sumerian empire 
under Ur III for a brief period reached to the Mediterranean Sea, 
Assyria, and Persia. However, according to the same authority, state 
trade prevailed again under the last Larsa ruler, Rim-Sin, 101 under 
the Babylonian king, Hammurabi, 102 who defeated him, and for over 
four centuries under the Kassites. 103 

In these compact hydraulic societies the "dense" bureaucratic ap- 
paratus obviously acted like a powerful hydraulic sponge, whose 
capacity to absorb vital functions of industry and trade was superior, 
other conditions being equal, to that of less compact hydraulic com- 



But such compact and self-perpetuating simple hydraulic societies 
are not too numerous. In many hydraulic civilizations the agromana- 
gerial apparatus state, while keeping the bulk of the cultivable land 
from becoming private property, did not so seriously restrict the 
growth of nongovernmental, property-based, and professional handi- 
craft and commerce. 

i. Occurrences 

a. Pre-Conquest Meso-America 

The rise of independent professional artisans and merchants in 
Aztec Mexico contrasts illuminatingly with conditions in Inca Peru. 
A complete lack of transport animals handicapped the inhabitants 
of Meso-America; but this deficiency was largely compensated for 
by a number of other ecological advantages. The terrain was far more 
suitable for interterritorial communication; navigable lakes, rivers, 
and an extended and approachable coast stimulated the circulation 
of goods by boat. The Sumerians enjoyed similar advantages; and we 
should not be surprised to learn that like them the Aztecs and their 
predecessors, the Toltecs, had private professional merchants and 
carried on an extensive international commerce. 1 These conditions 
also promoted a technical and regional division of industrial labor. 
But neither the city states nor the larger territorial units of pre- 
Conquest Mexico were as hydraulically compact as were their 
Sumerian counterparts. Thus the professional artisans and merchants 
of Mexico were not equally dependent upon the hydraulic state. Their 
plots of land were allotted by the calpulli, local and stratified units 
that possessed a limited autonomy; 2 and apparently neither group 
rendered extended labor services. Except for references to houses in 
which females were assembled, we have little evidence for govern- 
ment workshops. 6 According to Zurita and other early sources, the 
artisans rendered no corvee labor but paid over part of their produce 

a. According to Torquemada, houses with females, "nuns," were "widespread" 
(Torquemada, 1943, II: 189, 191). Diaz, who observed traditional Aztec society before 
it disintegrated, asserts that there were "nunneries" in a number of Central American 
countries. In Mexico proper he knew of only one, in the capital (Diaz, 1944, I: 349 ff.). 

b. Diaz (1944, I: 346) mentions government-managed bakeries. Sahagun (1938, III: 
75) speaks of persons who made shoes for the lords. Was the work in the government 
shops performed by serving men, who, while hereditary members of the calpulli, 
worked exclusively for the sovereign? (Monzon, 1949: 41.) Is this what Torquemada 
(1943, II: 488) had in mind when he said that certain work was done by artisans 

CHAPTER 7, G 259 

as tax. 3 Except for the time they spent in tilling their fields, the many 
Mexican craftsmen 4 seem to have deployed their special skills for 
their own account, preparing articles to be sold at the markets that 
were held in the large communities. 5 

The small traders were probably as independent as they were insig- 
nificant/' But the big interterritorial merchants, the pochteca, were 
close to the governmental apparatus. Permitted to rent out their 
plots of land G and to render tax instead of labor service, 7 the pochteca 
could engage in full-time commerce. They served the government 
as diplomats 8 and spies. 9 Occasionally they conducted military cam- 
paigns on behalf of their sovereign. 10 Tezozomoc says that the king's 
own brothers and uncles were pochteca. 11 

Manifestly, these big merchants were part of the ruling class. 12 But 
they were not commercial officials. Being rich, they operated with 
their own funds, and essentially, it would seem, for their own account. 
They might also collect taxes for the government, 13 and at such 
times, they were bureaucratic capitalists in the narrow sense of the 
term. However, this was no universal practice, for we know that 
as a rule the taxes were levied by full-time officials. 

And there is still less evidence that the Mexican pochteca and/or 
their aides traded largely on order of the ruler and the temples, as 
did the Sumerian damkar. Thus, however close the pochteca's as- 
sociations with the "lords" may have been socially and politically, 
professionally they do not seem to have been part of the state ap- 
paratus. It is for this reason and because of the independence of the 
artisans that we view Aztec Mexico as a semicomplex hydraulic society. 

The exact position of the Maya artisans is not easily determined. 
Clearly they were given fields, milpa/* and contrary to practice in 
Aztec Mexico they seem to have received allotments, not from semi- 
autonomous heads of the calpulli but from regional representatives 
of the central government. 15 The Maya commoners who built 
"houses" for the "lords" may well have included artisans; but the 
records are not articulate on this point. They are even less articulate 
on government-managed workshops, which as in Mexico were prob- 
ably not absent altogether. But as in Mexico, the Maya craftsmen 
probably produced and traded mainly for their own account. 16 

Lacking a comprehensive agromanagerial officialdom, the Maya 
rulers did not maintain an elaborate state trade. Some "rich" men 
were members of the governing class, 17 but it is doubtful whether the 

"for the lords"? Or are we faced with residual forms of an industrial corvee, which, 
although still invoked, had ceased to be institutionally relevant? 

c. Apparently they dealt in foods, cloth, and cacao on a modest scale and for a 
lowly clientele (Sahagun, 1938, III: 40, 53, 77). 


big Maya merchants in their entirety were as close socially to the 
secular and priestly leaders as were the pochteca. According to Landa, 
men of wealth lived near the "lords" and priests, but not in the 
same quarter. 18 Could it be that the crystallization of a property- 
based and nongovernmental group of professional merchants had 
advanced further in the hydraulically marginal lowlands of Yucatan 
than in the hydraulic core of Mexico? 

b. India, China, the Near East 

In India semicomplex patterns of hydraulic property and society 
prevailed throughout the greater part of its recorded history. In 
China and the Near East simple patterns of property yielded to more 
complex configurations and with differing results. China operated 
on a semicomplex level at least twice, once during the last centuries 
of the Chou period and again from the later part of the 5th century 
to the 8th century a.d. In the Near East complex patterns of property 
possibly prevailed only during a certain phase of Roman rule, 
whereas semicomplex configurations were prominent both before and 
after that time. 

Thus varying forms of semicomplex hydraulic property and society 
prevailed in India almost from the dawn of written history to the 
19th century, in China altogether for some five hundred years, and 
in the Near East for two long periods covering two thousand years 
or more. 

c. Byzantium and Russia 

I n Byzantine society there was no lack of private craftsmen and mer- 
chants. As a matter of fact, Byzantine trade was both comprehensive 
and flourishing during the middle and later part of the first millen- 
nium. 19 But the Byzantine artisans and merchants no longer had the 
freedom of action that their predecessors had enjoyed in the Greek 
cities of Western Asia or in Rome prior to the victory of bureaucratic 
absolutism. Administrative and fiscal restrictions burdened the crafts- 
men and traders of Byzantium until the 1 1 th century, 20 pressing them 
into a peculiarly crippled variant of a semicomplex pattern of hy- 
draulic property. 

In post-Mongol Russia private property in land evolved unevenly 
and as far as the peasants were concerned, very late. Professional and 
free handicraft recovered slowly from the setbacks instituted under 
the Mongol yoke. Commerce offered much greater opportunities to 
those who controlled it, and the masters of the Muscovite apparatus 
state were eager to manipulate it either directly, through trading 

CHAPTER 7, G 26l 

officials, or indirectly, through commercial agents. In the sphere of 
domestic trade government functionaries first purchased wax, honey, 
and other items, "taking them at smal prices what themselves list, 
and selling them againe at an excessive rate to their own marchants, 
and to marchants strangers. If they refuse to buy them, then to force 
them unto it." 21 The government also sold goods that it received 
as taxes or tributes and obviously with a similar disregard for the 
buyer, for such goods were "forced upon the marchants to be bought 
by them at the emperours price, whether they will nor no." d 

Foreign merchants too had to submit to government regulations. 
Once inside the Russian realm they had to display all their com- 
modities before the officials, who "put a value on them"; 22 and they 
could not trade with private individuals before the Tsar was given 
an opportunity to buy what he wanted. 23 

But the Muscovite state was unable to manage the bulk of all 
large-scale circulation as did the regimes of Pharaonic Egypt or Inca 
Peru. The Tsar comprehensively employed the services of a numbei 
of rich merchants, particularly the gosti. These bureaucratic capitalists, 
who collected taxes and custom fees for the government, 2 * usually 
acted as the Tsar's commercial councilors and agents. 25 

Outside the government trade proper, commerce was carried on, 
among others, by the pomeshchiki. These holders of office land sold 
the surplus grain and other surplus products of their estates for their 
own account, 26 thus constituting a group of bureaucratic capitalists 
sui generis. The monasteries, which were linked and subordinated 
to the state, also engaged in commercial transactions, not infrequently 
on a large scale. 27 

All this did not leave much room for the operations of professional 
and independent trade. The gosti and a small number of other 
privileged merchants controlled a large segment of the market, 28 
seeing to it that "nowhere free commerce be permitted." 2B Such 
at least was the opinion of the ordinary merchants, who played a 
decidedly inferior role and hated the gosti bitterly. 30 

Privileged merchants of the Muscovite period could amass great 
wealth, but neither this wealth nor their semi-official position pro- 
tected them against the confiscatory actions of their despotic masters. 
Fletcher reports a case in which three brothers of unusual energy 
and daring built up a thriving trade that yielded them "300,000 
rubbels in money, besides landes, cattels, and other commodities." 
Fletcher ascribes this initial success partly to the fact that the brothers 
lived more than a thousand miles from Moscow. For a while they 

d. The government profited particularly from the quasimonopolistic sale of furs, 
grain, and wood (Fletcher, 1856; 57 ff.). 


stood well with the authorities, who charged them with the adminis- 
tration of certain customs along the Siberian border. The Tsar was 
"content to use their purse, till such time as they got ground in 
Siberia." Finally, however, the government took away their fortune 
"by pieces, sometimes 20,000 rubbels at a time, sometime more; till 
in the end their sonnes that now are, are well eased of their stocke, 
and have but small parte of their fathers substance: the rest being 
drawen all into the emperours treasurie." 31 

Private property and property-based enterprise suffered immensely 
from this ruthless policy. "The great oppression over the poore 
commons," so Fletcher, 

maketh them to have no courage in following their trades: for 
that the more they have the more daunger they are in, not onely 
of their goods but of their lives also. And if they have any 
thing, they conceale it all they can, sometimes conveying it 
into monasteries, sometimes hiding it under the ground and 
in woods, as men are woont to doo where they are in feare of 
forreine invasion. ... I have seene them sometimes when they 
have layed open their commodities for a liking ... to look 
still behind them and towards every doore: as men in some 
feare, that looked to be set upon and surprised by some enimie. 32 

Under such conditions most of the commoners preferred immediate 
satisfaction to long-range planning: "This maketh the people (though 
otherwise hardened to beare any toile) to give themselves much to 
idlenes and drinking: as passing for no more then from hand to 
mouth." 33 It is difficult to find a more colorful and more depressing 
picture of private mobile property under the conditions of a crippled 
semicomplex Oriental society. 

2. How Powerful Could the Representatives of 
Private Mobile and Active Property Become in 
Semicomplex Hydraulic Societies? 

How much power may the potentially wealthiest representatives of 
mobile property, the big merchants, wield in semicomplex hydraulic 
societies? Can they ever dominate, or run, an absolutist government? 
Wealthy merchants certainly may control absolutist governments; 
and this may be the case even in commonwealths that contain ele- 
ments of hydraulic statecraft. Elements. As long as such governments 
fail to keep private property legally and economically weak, so long 
will the patterns of property and power remain hydraulically sub- 
marginal. This is always so when the interests of private property 
dominate the society; and it is so even when large hydraulic enter- 

CHAPTER 7, G 263 

prises and/or quasi-Oriental devices of political control are present. 
The city state of Venice built enormous protective water works, but 
Venice remained a nonhydraulic aristocratic republic, in which big 
commercial property gained a maximum of strength and security. 

Carthaginian society in the 4th and 3d centuries B.C. included a 
number of Oriental institutions. The Carthaginians certainly knew 
irrigation agriculture. 34 Their government was strong enough to tax 
the Lybian peasants of their agrarian hinterland/ To the disgust of 
their Roman enemies, they invoked the symbol of total submission, 
prostration, not only before their gods "as is the custom with other 
men," but also before their fellow men. 35 But as we have seen in 
Japan, irrigation techniques and prostration may occur also at the 
submarginal fringe of the hydraulic world; and in Carthage commer- 
cial interests were manifestly paramount f and private property was 
the key means for attaining high political office. 5 ' On the basis of our 
present knowledge we may therefore say that at least at the time of 
Aristotle the rich merchants probably dominated Carthaginian so- 
ciety and that similar submarginal configurations in all likelihood 
emerged in a number of other places, particularly — although not nec- 
essarily — at the geographical fringe of the hydraulic world. 

In independent commonwealths based on commerce, rich mer- 
chants — who may also be big landowners — can certainly achieve so- 
cial and political prominence. But while recognizing this possibility, 
we must ask: how much power can the representatives of independent 
commercial property wield in semicomplex Oriental societies? 

a. Miscellaneous Developments 

Under semicomplex conditions of property, the bulk of cultivable 
land is not owned privately; the big merchants must therefore derive 
their societal strength primarily from their mobile wealth. In a num- 
ber of cases their combined wealth was enormous; but even under ra- 
tional despots, such as the kings of Babylonia, commercial property 
generally remained subject to fragmenting laws of inheritance, to 
comprehensive taxation, and, insofar as transportation was concerned, 
not infrequently also to government regulation of oxen, carts, and 

e. Gsell assumes that normally the government claimed 25 per cent of the crops 
as tax. Polybius (1.78.2) shows that in emergencies as much as 50 per cent might be 
collected (Gsell, HA, II: 303). 
/. Meyer (GA, III: 644) calls the Carthaginian government a "commercial autocracy." 
g. Aristotle, Politics 2.11.1273a. Aristotle, who noted that in Carthage the greatest 
offices, such as those of kings and generals, were bought, considered this "a bad 
thing." "The law which allows this abuse makes wealth of more account than virtue." 
For an elaboration of these points see Gsell, HA, II: 235 ff. 


hired men. 38 It has been said before, and because of the importance 
of the issue it must be said again: the holders of active mobile prop- 
erty might organize in guilds, and often the state compelled them to 
do so; but neither the merchants nor the craft guilds were integrated 
in independent political machines on a local or national basis. 

The gentlemen traders of Aztec Mexico seem to have been content 
to act as a commercial appendage to the secular and religious rulers; 
and nothing is known regarding any attempts on their part to domi- 
nate Mexican society. The "rich" Maya, whose quarters were close to, 
but not identical with, those of the masters of the state, operated at 
the outer edge of the power system. Commoners, "apparently men of 
wealth or influence," sometimes "insinuated themselves into political 
positions considered to be above their station," but "the official hier- 
archy was purged from time to time of the pretenders and upstarts, 
who were not versed in the occult knowledge of the upper class." 37 

In the Old World, the marginal hydraulic societies of Byzantium 
and Russia differed greatly from Maya society, but their private trad- 
ers also failed to become politically dominant. In Byzantium the mer- 
chants, however wealthy they were individually, remained politically 
and socially restricted until the nth century. During the final phases 
of Byzantine history, the men of property who succeeded in paralyz- 
ing the absolutist apparatus were not merchants or artisans, but land- 

In Muscovite Russia the merchants were little more than economi- 
cally useful domestic animals; nor did the big merchants in China 
rise to political prominence when semicomplex patterns of property 
prevailed at the end of the Chou period and during the middle part 
of the first millennium a.d. 

b. Hindu India 

The corresponding developments in early India are particularly in- 
structive because the Aryan conquest was accomplished by a group 
that, although aware of the importance of irrigation canals, 38 empha- 
sized cattle wealth, trade, and traders. The Vedas speak respectfully 
of merchants.* In a hymn in the Atharva-Veda-Samhita merchants 
pray to the god Indra as "the merchant par excellence." S9 The great 
epics that were composed very much later 40 confirm the relatively 
high and influential position of the Vedic merchant in what Hop- 
kins calls "the Aryan state." 41 However, they leave no doubt that 

h. Grassmann, RV, I: 197; II: 113; cf. Banerjee, 1925: 155. Less esteemed, although 
equally prosperous, was the pani, a businessman who sought gain "either through 
trade or through usury" (Banerjee, 1925: 156). 

CHAPTER 7, G 265 

"in distinction from nobles and priests," the merchants, together with 
the Aryan peasants, belonged to "the people." 42 Thus whatever the 
status of the Aryan commoners, the Vaisyas, may have been in pre- 
historic times, in the Vedic era they were "oppressed by the princes." 
It was in this era — or even later in the subsequent Buddhist period 43 
— that professional associations of merchants began to appear.** 

Of course, the rise of such bodies proves nothing about their politi- 
cal independence. In simple Oriental societies — and often also un- 
der more complex conditions — the professional corporations are use- 
ful tools of government.. The epics voice the king's concern with the 
merchants, particularly in times of war and crisis; but the merchants' 
chief political importance may well have been derived from their 
possible conspiratorial value to enemy countries. 45 

There can be no doubt regarding the prospering of trade and 
traders during the Buddhist period; and there can be no doubt 
either regarding the social prominence of the government-attached 
chief merchants, the setthi. However, this does not justify the claim 
that the merchants, as a group, were able in the major centers of 
what was then Hindu India to normally and conspicuously influence 
— or control — the political decisions of their respective governments. 

These governments were not necessarily monarchies. In the home- 
land of Buddhism, northeast India, there were several republics, in 
which the ruler discussed public affairs in full and frequent assem- 
blies. 46 But the merchants were not included in these bodies. The 
meager information we have on eight of the ten republics listed by 
T. W. Rhys-Davids 47 shows all of them to have been dominated by 
members of the warrior caste, Kshatriyas. 48 Buddha considered their 
assemblies an ancient institution; *° and it may well be that the pat- 
terns of Aryan society * persisted somewhat longer in the northeast- 
ern area, in which hydraulic action, although highly advantageous, 
was not so crucial as in the more arid western parts of the north In- 
dian plains.*' However, irrigation agriculture and hydraulic enter- 
prises were by no means absent in the northeast; 50 and the aristo- 
cratic republics clearly moved toward a monarchical form of power 51 
which was already widespread in the days of Buddha k and which, 
after a transitional period of turmoil and conquest, came to prevail 
throughout the heartlands of Aryan culture. 52 

1. For the original role of an aristocracy of warriors see Hopkins, 1888: 73; Keith, 
192a: 98. 

/'. Cf. Stamp, 1938: 299 ff. Oldenburg (1915: 284) regrets that the studies of Vedic and 
Buddhist India have neglected the solidly Brahmin development in the west and the 
great susceptibility of the east to the anti-Brahmin movement of Buddhism. 

k. For the despotic character of these Indian monarchies see Law, 1941: 169 ff. Cf. 
Fick, 1920: 105 ff. 


In the restless and changing Indian society of this important pe- 
riod many governments availed themselves of the services of a setthi. 
Apparently a man of means, 53 the setthi often advised and aided the 
ruler in economic matters. 54 His position, though not that of an offi- 
cial, 55 was distinguished and hereditary, 56 vacancies being filled by 
the king. 57 

The term setthi means "best, chief." 58 Manifestly he was a "repre- 
sentative of the commercial community," 59 but it is most important 
to note that he did not operate as the constitutionally established 
spokesman of organized merchant power. Nor does he seem to have 
been regularly — or primarily — concerned with guild affairs. His title 
"may possibly imply headship over some class of industry or trad- 
ing"; 60 and a famous setthi mentioned in the Jataka tales apparently 
"had some authority over his fellow-traders." 61 But this authority, 
even if real, was rooted in a body whose organizational effectiveness 
has not yet been clearly established. In Buddhist and post-Buddhist 
India there certainly were merchant corporations, but C. A. F. Rhys- 
Davids warns against over-estimating the degree to which the traders 
were syndicalized. 62 To repeat her conclusion: "There is ... no 
instance as yet produced from early Buddhist documents pointing to 
any corporate organisation of the nature of a gild or Hansa league." 63 

All this does not preclude the political prominence of merchants 
in some Orientally submarginal cities or city states of classical India; 
but it stresses the need for a most careful examination of the sources 
adduced to prove such prominence. 

Hopkins, the well-known Sanscritist, cites a Nepalese legend of 
the 3d or 4th century a.d. as offering particularly valuable data on 
the political power of a merchant guild." 1 In his opinion this legend 
"records that Thana was ruled by a strong merchant guild." 64 Turn- 
ing to the Bombay Gazetteer which Hopkins consulted, 65 we find that 
it makes a significantly more limited claim: "A strong merchant 
guild ruled the trade of the city." 66 The city in question is Sopara, 
one of several settlements located on the coast of Thana, 67 south of 
modern Bombay. Turning to the legend itself, we find that the mer- 
chants in question, far from controlling the government of the city, 
did not even control its trade. A single powerful outsider prevailed 
over the "500" merchants who were trying to corner the market, and 
he did so after both parties were summoned to appear before the 
king, who manifestly was the undisputed ruler of the city and the 
merchants. 88 

The Indian development is instructive in several respects. The 

m. "Later literature down to our own time contains frequent reference to such 
bodies, but no thorough treatment of them is to be found" (Hopkins, 1902: 175). 

CHAPTER 7, G 267 

Kshatriya republics show that hydraulic regimes need not be mo- 
narchic; but their final phases also underline the tendency toward 
a concentration of power that inheres in such regimes. The fate of 
the merchants is equally worth noting. During the formative days 
of the Aryan conquest society, traders enjoyed considerable social 
prestige. But subsequently their position deteriorated, and this hap- 
pened despite the fact that they were tightly organized. 

c. Ancient Mesopotamia 

Were the merchants more successful in the great Western Asiatic 
cradle of Oriental trade, ancient Lower Mesopotamia? Sumerian 
legends speak of elders and assembly-like gatherings, which the 
legendary king, Gilgamesh, consulted before making decisions. 69 
What do these tales mean? Boas has convincingly argued that myths 
contain fictitious as well as realistic features and that realistic ele- 
ments may be exaggerated or transformed into their opposites. 70 
There may very well have been proto-Sumerian assemblies similar to 
the warrior assemblies of the Aryan conquest republics in northeast 
India. Kramer assumes the existence of a military aristocracy during 
the formative period of prehistoric Sumer. 71 But whatever the insti- 
tutional quality of these legendary assemblies may have been, no 
such gatherings dominated the Sumerian city states when they 
emerged in the light of recorded history. To quote Jacobsen: "The 
political development in early historical times seems to lie under the 
spell of one controlling idea: concentration of political power in 
as few hands as possible." 72 In each of the early Mesopotamian city 
states "one individual, the ruler, united in his hands the chief politi- 
cal powers: legislative, judiciary, and executive." 78 In each of them 
the king handled the despotic state apparatus through the agency of 
an effective secular and priestly bureaucracy, "the court and temple 
administrators and intellectuals," as Kramer calls the new core of 
the "ruling caste." 74 

Significantly there are few, if any, traces of assemblies in the sim- 
ple hydraulic society of historical Sumer. With regard to Babylonia 
the situation is otherwise. Babylonian inscriptions refer to assem- 
blies, to elders, and — in the same context — to merchants. Could it be 
that the growth of Babylonian trade also increased the power of its 
representatives, the big merchants? 

The possible extent, and the limitations, of merchant power are 
indicated by the Assyrian merchant colonies, which flourished in 
Cappadocia during the earlier part of the second millennium B.C. 
These Assyrian settlements were established in an area which, al- 


though lacking political unity, 75 comprised a number of territorial 

The Assyrian traders who settled far to the north of their home- 
lands did not dwell inside the Cappadocian towns. The walled sec- 
tions were reserved for the native population and for the palaces of 
the ruler. 78 Moreover, the local authorities 77 inspected the trader's 
commodities in the palace and also, it seems, had a first claim on any 
goods they wanted to sell. 78 The presence of such local authorities 
did not mean that the colonies were independent of the Assyrian 
metropolis. In the end it was Assur that decided legal cases and that 
had the power to impose taxes: 78 "The authorities of Assur and ulti- 
mately the king were therefore the superiors of the Assyrian authori- 
ties in the commercial centers." 80 

Within this over-all frame the colonies dealt with their judicial 
matters in "a general assembly of all colonists," 81 the karum; and 
this body also settled other communal problems. 82 Evidently the 
members of these Assyrian trade colonies enjoyed a greater autonomy 
than did the merchants of Assyria or Sumer, or — after the close of 
the Sumerian period — Babylonia; but they did not dominate the 
Cappadocian towns, nor were they politically independent in their 
own quarters. 

Babylonian absolutism, like that of Sumer, was rooted in a com- 
pact agromanagerial economy; and private property probably played 
a secondary role in agriculture as well as in commerce." In any case, 
no serious institutional analyst claims that the assemblies, and 
through them the merchants, controlled the Babylonian govern- 
ment. The king and his men dominated the administration, the 
army, and the fiscal system. The king was also the lawgiver. Further- 
more, he and his functionaries were strategically situated in the ju- 
diciary. At the king's service "judges of the king" ruled according 
to the "legal practice of the king." 83 But the royal judges, who fre- 
quently combined administrative, military, and legal activities, 84 re- 

n. Probably. The reasons for the second part of our assumption have been given 
above; the reasons for the first will be given below when we discuss the extent of 
private landownership. Dr. Isaac Mendelsohn, in a personal communication and on 
the basis of an independent examination of the inscriptions, believes that in both 
spheres of Babylonian economy private property was more extended than the com- 
bined property of the state and the temples. No doubt the facts of the matter have 
to be decided by the period specialists; and our tentative classification of Babylonian 
society is therefore open to whatever adjustments future research may postulate. 
But assuming for the sake of the argument that the private property sector exceeded 
the public sector, there is still no need to change our evaluation of the subordinate 
political position of the Babylonian merchants. In the same personal communication, 
Dr. Mendelsohn rejects an interpretation of Babylonian society as democratic. 

CHAPTER 7, G 269 

lied for the settlement of local issues heavily on local assemblies. 
These bodies dealt primarily with legal matters. 85 Operating under 
the king's control, they constituted "a kind of civil jury." ° 

The members of these assemblies were "elders," "notables," "mer- 
chants" (under a head merchant), and "men of the gate." 86 Accord- 
ing to Cuq, these designations refer to separate groups that acted 
either alone or in combination. 87 Whether Cuq's interpretation is 
correct or not and whatever the terms "elders," "notables," or "men 
of the gate" may mean, for our present purpose it is sufficient to 
know that the assemblies were essentially judicial bodies and that 
among their members there were merchants headed by an akil 

In early Babylonia the akil tamgari seems to have been the direc- 
tor of the Department of Commerce or the Department of Finance, 
and as such the chief of the fiscal bureaucracy. 88 He headed the ordi- 
nary merchants, who undertook commercial expeditions, "at times 
exclusively in the interest of the crown." p He thus was a prominent 
official through whom the absolutist regime exerted control over the 
country's traders. 

Occasionally an assembly dealt with issues that concerned a whole 
town; and its merchant members would therefore be participating 
in matters of considerable local importance. However, since the as- 
sembly was presided over by a royal governor or town prefect and 
since it acted essentially as a civil jury, it certainly did not control 
the town government; and the merchants, who were under the au- 
thority of the akil tamgari, were not free to control even their own 
professional spheres, the country's trade. 

d. Conclusions 

The lessons of all this are obvious. Powerful groups of rich mer- 
chants may control the government of their commonwealth; and this 
may happen even in communities that fulfill substantial hydraulic 
functions. But as far as we know, such developments did not result 
in anything that can be called the rule of hydraulic merchants. The 
great merchants of Venice operated in a societal setting in which hy- 

o. Cuq, 1929: 361. Occasionally they also handled political crimes, but the case cited 
by Jacobsen involves no deeds, but words only: "seditious utterances" (Jacobsen, 
1943: 164). 

p. Kriickmann, 1932: 446. Was the head merchant of the king, rab tamqar sa iarri, 
who is mentioned in the Neo-Babylonian inscriptions, the successor of the akil tamqari? 
His activities were obscure. Ebeling (1933: 454) places him among the "high 
officials," adding that he "probably conducted commercial and monetary transactions 
for the king." 


draulic institutions were submarginal. And Carthage, although cer- 
tainly more hydraulic than Venice, may well have belonged, either 
from the start or eventually, to the submarginal zone of the hy- 
draulic world. 

Carthage-like or Venice-like commercial commonwealths flour- 
ished in considerable numbers at the geographical fringe of hy- 
draulic society; and there is no reason why such commonwealths 
should not have constituted independent heterogeneous enclaves 
also within certain zones of the hydraulic world. We therefore do 
not reject Max Weber's assumption that independent commercial 
communities may have flourished in Buddhist India. 89 But the evi- 
dence adduced is not conclusive; and in a number of cases reexamina- 
tion reveals that the position of the merchants is far from being 
politically dominant. 

Further inquiries into the political role of merchants in institu- 
tionally peripheral regions will certainly deepen our insight into the 
diversities that exist within the margin and the submargin of the 
hydraulic world. They may also shed more light on the limitations 
of mobile private property even in those hydraulic societies in which 
private-property-based commerce became more important than gov- 
ernment-managed and government-attached trade. 


1. Hydraulic Landlordism, Past and Present 

The limitations of immobile property in hydraulic society are 
equally significant — and equally misunderstood. The institutional 
pioneers who viewed the despotic state as the only major landowner 
tended to neglect the problem of private landownership altogether. 
Modern observers, who have noted the paralyzing influence of ab- 
sentee landlordism in the Orient, are inclined to treat as a basic fea- 
ture of hydraulic society what in many cases is only a feature of 
hydraulic society in transition. And they are quick to interpret in 
terms of past (feudal) or present (capitalist) Western institutions 
what is actually a specific Oriental development. 

a. To mention just one key issue: the establishment of private peasant land by 
means of a thoroughgoing land reform has one meaning when it is undertaken by 
the separate forces of a relatively decentralized postfeudal or industrial society 
and quite another when it is undertaken by the government-controlled forces of a 
disintegrating hydraulic order or, for that matter, by a totalitarian state of the Soviet 
type. Major changes in the system of land tenure that occurred in modern Japan, 
in Russia under the Tsars or under the Bolsheviks, in Nehru's India, or in Communist 

CHAPTER 7, H 271 

More will be said on this subject in our concluding chapter. In 
the present context we are concerned essentially with the roots of 
the modern development: the extent and peculiarities of private 
landownership prior to the dissolution of hydraulic society. 

2. Government-controlled and Private Land in 
Hydraulic Society 

The extent and the peculiarities of private land in hydraulic society 
can be properly viewed only when we remember the extent and the 
peculiarities of hydraulic state power. In the majority of all hydraulic 
societies the despotic regime kept private land in a quantitatively 
subordinate position. In all hydraulic societies the despotic regime 
limited the freedom of the private land it permitted to exist. 

a. Types of Government-controlled Land 

In order to establish the extent of private land we have to clarify 
the extent of government-controlled land. This last comprises three 
main types: (1) government-managed land, (2) government-regulated 
land, and (3) government-assigned land. 

All land that is kept by government measures from being alienated 
either to or by private landowners is regulated land in the broad 
sense of the term, and in this sense all government land is regulated 
land. In a narrow sense, the term "regulated land" will be applied 
essentially to that part of government-controlled land that is man- 
aged not by the government but by possessors, who work for, or pay 
tax or rent to, the government. The term "government-managed 
land" will be applied to land that is farmed under the direction of 
government functionaries and for the immediate and exclusive bene- 
fit of the government. The term "assigned land" will be applied to 
land that is temporarily, or indefinitely, assigned to officials (office 
land), to representatives of the dominant religion (sacred or temple 
land), or to some distinguished persons who do not, in return, fulfill 
any special secular or religious functions (sinecure land). 

i. government-managed land 

Government-managed "public" land was never more than a mi- 
nor part of all regulated land, since the peasants who cultivated the 
"public" fields also needed land for their own support. Above a cer- 
tain agronomical level and except in some strategically important 

China are frequently treated as if they were more or less identical, though in their 
societal substance and effect they are entirely different phenomena. 


regions, the hydraulic state preferred the payment of a land tax from 
the individually cultivated fields to products from public fields. 

Imperial China, although favoring private ownership of land, 
maintained farm colonies for the support of the army, primarily in 
border areas, but at times also in critical inland areas: at places that 
were being "pacified" and along vital lines of communication. The 
tilling in these colonies was done either by soldiers (in which case 
they were generally called "garrison fields," t'un-t'ien) or by civilians 
(in which case they were frequently called "camp fields," yin-t'ien). 
The two types of fields together occasionally comprised as much as 
one-tenth of all cultivable land, but in most dynasties the fraction 
was much smaller. 

Apart from military colonies, there were government domains for 
the growing of special crops, and parks and gardens for the rulers' 
pleasure. These secluded retreats were often built with corvee labor, 
but usually they were cared for by professional cultivators, palace 
laborers, and slaves 6 — that is, they were government-managed. But 
while remarkable in this respect, they were spatially insignificant. 
They were tiny islands in a sea of peasant farms, whose occupiers or 
owners supported the government not by their labor or public fields 
but by their tax payments. 


The most important type of all government-controlled land is per- 
haps the least clearly defined: peasant land which is neither managed 
by government officials nor assigned to groups of grantees, nor owned 
by the cultivators. This type of land cannot be simply equated with 
the land of village communities, since not all peasants who possess 
regulated land live in integrated village communities — that is, in 
communities that distribute and redistribute the land. Nor are all 
village communities under the control of the government. 

Regulated peasant land, in terms of the present inquiry, is land 
that a holder cannot alienate freely. Often, and particularly when 
the land is periodically redistributed, a holder may be allowed to 
lease it to other villagers, but he cannot sell it. d In other cases he 

b. Royal or imperial gardens and parks have been described by many authors. For 
the Mexican lake area see Ixtlilxochitl, OH, II: 209 ff.; for Pharaonic Egypt, Erman 
and Ranke, 1923: 206 ff.; for ancient Mesopotamia, Meissner, BA, I: 201, 292; Contenau, 
!95° : 53 ff -; for the Islamic Near East, Mez, 1922: 362 ff.; for Muslim Spain, Levi- 
Provencal, 1932: 223; for India, Jatakam: passim and Smith, 1926: 402 ff.; for Chou 
China, Legge, CC, II: 127 ff. 

c. This was customary among the calpulli members of Aztec Mexico. See Zurita, 
1941: 88; Monzon, 1949: 39. 

d. For an elaborate description of the regulated village community in Tsarist Russia, 
the obshchina or mir, see Haxthausen, SR, I: 129 and passim. 

CHAPTER 7, H 273 

may sell it but only to other villagers — that is, to fellow peasants. In 
Byzantium earlier directives were restored and reenforced in 922 
by a law which permitted the peasants to sell land to the following 
groups and in this order: (1) co-possessing relatives, (2) other co- 
possessors, (3) persons whose land was adjacent to the land to be 
sold, (4) neighbors who shared the seller's fiscal responsibility, and 
(5) other neighbors. 1 These regulations made it impossible for a 
landlord to purchase peasant land except in villages where he was 
already an owner. 2 As long as they worked, they protected the bulk 
of the peasant land from falling prey to the expanding forces of land- 

Similar principles were employed in Hindu e and Muslim India. 
Bulwarked by the law-enforcing powers of the state, the Indian village 
community "protected small farming against the invasion of capital- 
istic interests," and it did so "by maintaining [for the villagers] the 
rights of entail, pre-emption, and preoccupation." s 

The cases of Byzantium and India, which could be supplemented 
by data from other civilizations, demonstrate the negative effects of 
regulated land on the growth of private landownership. Wherever 
the Orientally despotic state insisted on keeping the bulk of all land 
regulated, private ownership of land was kept in a secondary and not 
infrequently in an irrelevant position. 


The despotic regime that is able to regulate all or a large part of the 
land is also able to assign portions of it to any individual or group of 
individuals. Such land assignments may differ in purpose and dura- 
tion, but usually the two aspects interlock. Persons who serve the 
government may hold their office land for life or even hereditarily. 
Others may hold their offices only for a short term; in such cases 
tenure over their office land is equally brief. Serving men who fulfill 
military functions are particularly apt both to obtain and to lose 
their office land suddenly. 

Land grants made to those who serve the gods are more stable. En- 
during religious organizations, such as temples and mosques, are al- 
most always permitted to retain their grants indefinitely. 

Sinecure land is given for a variety of reasons to a variety of per- 
sons. The grantees may be so distinguished because of their meritori- 

e. See Appadorai, 1936, I: 133 ft. The alienability of land has been seen as a sign 
of ownership, whereas it may merely be indicative of a flexible form of possession. 
Jolly's (1896: 94) interpretation makes allowance both for the (externally) regulated 
and the (internally) fluid conditions of village land. He assumes "that generally the 
villages were shut off from the outer world, but that within the individual villages 
there existed private property of land." 


ous acts or merely because they are the ruler's relatives, friends, or 
favorites/ In all instances the land is assigned unconditionally. The 
grantees do not render service for the revenues which the sinecure 
land yields. This is also true for the holders of pension land. But 
whoever the beneficiary may be, the government remains the master 
of the assigned land. 

Sacred (temple) land is usually supervised and/or managed by 
secular government officials. This has been established for Pharaonic 
Egypt,* for Ptolemaic Egypt, 5 for Babylonia, 6 and of course, for pre- 
Conquest Peru and Mexico. In the Islamic world, direct or indirect 
state control over the various types of religious property persisted, 
with many modifications in detail, until recent times. 7 

Control over office land is guaranteed by the government's opera- 
tional control over the landholders. A normally functioning despotic 
regime determines the fate of its serving men and the lands allotted 
to them. When, at the close of the Chou period, the chancellor of 
the state of Ch'in made merit rather than inheritance the essential 
basis for office,- 9 he met with no conspicuous resistance; and through- 
out late Chou China the decrease of areas administered by holders 
of office land 8 were accepted with equal meekness. No organized 
group of "barons" rose against the imperial unifier of China when he 
finally and decisively discarded the office land system in its entirety. 
Nor did Akbar's decision to substitute in large part salaries for office 
land 9 meet with any greater challenge. Akbar went far, but not so 
far as the Turkish Sultan Suleiman, who spectacularly demonstrated 
that a well-functioning despotism could abolish office land as easily 
as it could create it. 10 

Sinecure land might be given without any limitation as to time. 
In this case possession might come to an end when the ruling dynasty 
fell. In Pharaonic Egypt this seems to have been the rule; " and it is 
not unlikely that the land grants in ancient Peru would have suffered 
the same fate if the Inca regime had been replaced by other native 
rulers. Often sinecure land was intended to support the recipient as 
long as he lived, but the grantor's death might terminate the assign- 
ment earlier. The land grants of ancient Hawaii were apparently so 
conditioned. 12 

f. Cf. Jatakam, I: 56 (grant given to the king's barber); II: 193 (to a Brahmin), 270 
(to a princess), 457 ff. (to a princess); IV: 1 x6 (to a Brahmin), 309 (reward for finding a 
precious antelope), 415 (to a princess), 480 (reward for singing a special verse); V: 21 
(reward for useful advice), 35 (to ascetics), 45 (to a hunter), 374 (to a hunter); VI: 135 
(to a barber), 355 (to the king's brother or son), 422 (to a setthi), 438 (to good advisors), 
447 (to an advisor). Cf. ibid., I: 362 ff., 424, 462. 

g. Shih Chi 68.4a; Duyvendak, 1928: 15, 61. The "nobles" whom he restricted more 

CHAPTER 7, H 275 

b. Private Land 


Land that is government-managed, government-regulated, or gov- 
ernment-assigned is obviously not the property of private landowners; 
and it cannot be so viewed, even when possession is prolonged. Per- 
manency of possession is not enough (hereditary tenants also enjoy 
this privilege); nor is the right to alienate enough (holders of 
regulated land are sometimes permitted to alienate it within their 
social group). Only when the proprietor has the right both to hold 
his land indefinitely and to alienate it to persons outside his social 
group do we encounter what, in conformance with established usage, 
can be called full private landownership. 


The commoners and nobles of early Greece, Germany, Gaul, and 
England owned their land not because of the decision of an auto- 
cratic ruler but because of differentiations within a tribal society, 
which produced multiple patterns of private property and political 
leadership. In hydraulic society it was essentially the ruler and his 
functionaries who established private landholding by transferring to 
individual owners what was previously government-controlled land. 

Individuals usually became landowners through gifts or sale. Entire 
groups were made landowners by government decree. After a piece 
of land had been recognized as private property, it could, within 
government-set social limits, be transferred from one private owner 
to another. Large-scale conversions of regulated land into private 
land are relatively rare in the history of Oriental society. They seem 
to have occurred only where private-property-based handicraft and 
trade were well developed. 

c. Types of Landownership 


Who then are the potential owners of land in hydraulic society? In 
Oriental as in other agrarian societies the key figure in the basic 
subsistence economy is the peasant. We can therefore expect him to 
play an important role in the expanding sector of private land- 
ownership; and indeed in China the establishment of free private 
land involved the emergence of a large class of peasant owners. 

and more (ibid.: 27; Shih Chi 68.8b) are said to have hated him (Duyvendak, 1928: 23; 
Shih Chi 68.6b), but his measures led to no organized "baronial" rebellion. 



But the Chinese development is the exception rather than the rule. 
In the majority of all cases it is not the peasant owner but the non- 
peasant owner who first and prominently appears in the private land 
sector. Evidently the more complex a hydraulic society becomes, 
the greater the number of social groups that seek to be landed 
proprietors. But one group among them is outstanding: the civil 
and military functionaries of the government and their relatives, 
the bureaucratic gentry. 

Under simple conditions of property few others are rich enough 
to buy land. And even where there are wealthy merchants or 
traders, the bulk of the surplus, and consequently the bulk of the 
purchasing power, remains in the hands of the governing class. 
Furthermore, it is to members of the governing class that the ruler 
is most likely to make gifts of land. 

Bureaucratic landlordism therefore tends to appear in all types of 
hydraulic society, whatever their complexity. It prevails completely 
in those simple hydraulic societies in which private land is at all 
relevant. It is a significant feature in many semicomplex hydraulic 
societies. And it is crucial in complex hydraulic societies where 
privately owned land outweighs state-controlled land. 

Data on landed property in Pharaonic Egypt are vague even for 
the New Kingdom. 13 A few statements that are specific speak essen- 
tially of princes, viziers, and other members of the governing class 
as owners of private land. 14 

In Aztec Mexico private lands were held by the rulers, their 
officials, and some merchants. 15 In Hindu India the Brahmins did 
not, as was the case with priesthoods in many other hydraulic soci- 
eties, live on large and permanently granted temple lands. Conse- 
quently, in Hindu India land grants to individual Brahmins fulfilled 
a special function, and it is not surprising to find that they were 
numerous. Many of them carried only the right of possession, but 
a number of Brahmins seem to have owned land at least in the 
last phase of Hindu rule. 16 In Byzantine Egypt the "powerful ones" 
who had large estates were most frequently officials; 1T and this 
pattern is repeated in Islamic times. Among the persons who, during 
the Mamluk period, acquired private land, actual or former holders 
of office land were prominent. 18 In Ottoman Turkey some office 
lands became the private property of former holders. 19 

In Middle Byzantium functionaries were for a time forbidden 
without special imperial permission to purchase land while they 
held office. The restriction retarded the growth of bureaucratic land- 



ownership but did not prevent it. 20 In Tsarist Russia the edict of 
1762 converted the pomeshchiki, who had been possessors of office 
land, into landowners. In later imperial China government func- 
tionaries were forbidden to purchase land in the district in which 
they officiated. 21 Nothing was said regarding the purchase of land 
outside this area; and the evidence at hand suggests that among the 
owners of land officiating and nonofficiating members of the govern- 
ment class were outstanding. 


To be sure, members of other social groups also owned land, if they 
had the necessary means and if they were permitted to. In semi- 
complex and complex hydraulic societies rich merchants particularly 
were likely to acquire land; and information on Aztec Mexico, 22 
India, 23 and China shows clearly that they did so. Moreover, the 
measures invoked by the Han dynasty reveal both how well en- 
trenched this type of landlordism might become and how ruthlessly 
a ruling bureaucracy might combat it. 24 Of course, even persons of 
modest wealth might buy land. In traditional China persons from 
all walks of life owned small pieces of land 



Occasionally a nonpeasant owner of land, who for some reason 
or other was deprived of his occupation, might assure his support by 
personally turning to farming. h Generally, however, nonpeasant 
landowners left the tasks of cultivation to tenants. In many cases 
they were absentee landlords. 

In Medieval and post-Medieval Europe tenancy and absentee land- 
lordism were also widespread. However, many landlords personally 
managed their large estates (Giiter) or employed stewards for this 

The small incidence of large-scale farming in hydraulic society is 
due primarily to the high crop yield obtained by labor-intensive 
methods, which are in part required and in part stimulated by irriga- 
tion agriculture. 26 These methods provide extraordinary advantages 
for small-scale peasant farming on a family basis. The advantages are 
so striking that the dominant hydraulic "economic ethos" (Wirt- 
schaftsgesinnung) discouraged large-scale and "manorial" methods, 
even when they might have been profitably applied. 

The significance of this attitude for hydraulic society in transition 

h. For Brahmins who tilled their land either with or without the aid of farmhands, 
see Jatakam, II: 191 ff.; Ill: 179, 316; IV: 195, 334 ff.; V: 70. 


is obvious. The consolidation of landlordism in postfeudal Europe 
encouraged many owners of large farms to cultivate their land 
scientifically. The recent growth of landlordism in many hydraulic 
countries intensified the acquisitive zeal of the absentee landlords 
without increasing the rationality of tenant farming. 


An interesting variant of absentee landlordism appeared in Tsarist 
Russia. The pomeshchiki of Muscovite and post-Muscovite Russia 
were kept so busy rendering military or civil services that they 
could not pay much attention to farming, as did the landed nobles 
of England or Germany. In consequence, large-scale and scientific 
farming was extremely limited among landholding aristocrats in 
Russia prior to 1762, and despite some expansion it remained the 
exception long after this date. 

Baron Haxthausen, who made his famous study of rural Russia 
in the 1840's, was struck by the difference between landlords in 
Russia and in the rest of Europe. Although unaware of the peculi- 
arities of Oriental despotism, he clearly recognized that Russia's 
landowning aristocracy lacked a feudal tradition: 

The Russian, the Great-Russian nobility, is not a landed nobility 
[Landadel] now, nor in all probability was it ever one; it had 
no castles, it did not pass through a period of knighthood and 
[private] feuding. It always was a serving nobility, it always 
lived at the Courts of the Great Princes and the smaller princes 
and in the cities, rendering military, Court, or civil services. 
Those among them, who lived in the countryside, peacefully 
pursued agriculture; but in actuality they were either insignifi- 
cant or unfit. Even today the majority of the Great-Russian 
nobles have no rural residences, no [manorial] economies as we 
see them in the rest of Europe. All the land that belongs to the 
noble — cultivated land, meadows, forests — are left to the peasant 
village community that works it and pays the lord for it. Even 
if the lord owns, and lives in, a country house, he still does not 
have a [manorial] economy, but rather lives like a rentier. Most 
nobles have country houses, but they live in town and visit the 
country house only for weeks or months. This is the old Russian 
way of life of the aristocracy! 27 

The Russian nobles' peculiar detachment from the land they 
owned — together with the fragmenting law of inheritance — kept 
them from becoming "a real landed aristocracy," as Haxthausen 

CHAPTER 7, H 279 

knew it in Central and Western Europe. "I do not think that there 
is, in any major country in Europe, less stability of their land than 
in Great Russia." 28 

It is against this background that we must view the two great 
agrarian changes accomplished by the Tsarist bureaucracy in the 
second half of the 19th and the beginning of the 20th century: the 
emancipation of the serfs from their former landlords (in 1861) and 
Stolypin's reform (in 1908). In both cases resistance was great; but in 
both cases the new measures were introduced by members of the 
same governing class that comprised the bulk of all landlords. 


Absentee landlordism is quickly apparent, more quickly than the 
exact proprietary quality of a particular piece of land. How many 
of the land "grants" of Pharaonic Egypt or Buddhist India were 
given with the intention of establishing possession? How many with 
the intention of establishing ownership? The records often fail 
to provide definite information on these points. And even when they 
suggest the right of ownership — how secure was this right? Segre, 
in comparing the proprietary developments under Oriental ab- 
solutism and in classical Greece, concludes that "private property 
in a sense approaching to classical ownership could not exist as long 
as the king could exercise the power of withdrawing rights either 
to land or to liberties or change these terms at will." 29 

Brahmin property was believed to be safe from confiscation. But 
this did not prevent Hindu rulers from seizing Brahmin land for 
"treason," which the king's judiciary had no difficulty in establishing 
when it suited his purposes. 30 In Pharaonic Egypt private landowner- 
ship, although perhaps more extensive than in Hindu India, was 
equally insecure. Indeed it was "basically nothing but an exceptional 
transfer of royal prerogatives, a transfer which as a matter of princi- 
ple could be reversed at any time and which was often reversed when 
a new dynasty came into being." 31 In such cases it is manifestly hard 
to draw a sharp line between possession and ownership. 

Another difficulty arises from the fact that in certain hydraulic 
societies the right to alienate private property in land is spread 
unevenly. Nonpeasant landlords may be free to buy land from other 
landlords, whereas the peasants who live in a regulated rural order 
enjoy no corresponding right of alienation. In hydraulic society 
such mixed patterns create a major classificatory problem only when, 
as in Late Byzantium and in Russia after 1762, the land held by 


landlords comprises a large part (perhaps more than one-half) of all 
cultivated land. When this is the case, we can speak of an incipient 
pattern of complex hydraulic property and society. 

d. The Extent of Private Landownership in Various 
Subtypes of Hydraulic Society 

The categories of government and private land developed so far 
enable us to advance beyond our initial tentative position and to 
correlate with greater precision and fuller evidence the advance of 
mobile and immobile private property in various hydraulic civiliza- 
tions. Germs of private landownership were present even in hydraulic 
societies in which private-property- based industry and commerce were 
of little consequence, but they did not assume major dimensions. 
This confirms the validity of our concept of "simple" patterns of 
hydraulic property and society. In hydraulic civilizations with a 
substantial sector of mobile property and enterprise, private land- 
ownership frequently remained a secondary feature, and occasionally 
an insignificant one. This confirms the validity of our concept of 
semicomplex patterns of hydraulic property and society. Furthermore 
it confirms our contention concerning the relative scarcity of the 
complex configuration — a configuration in which immobile private 
property is as prominent in agriculture as, in its peculiar way and 
with its peculiar limitations, mobile property is prominent in in- 
dustry and trade. 

On the basis of these results, we shall contemplate briefly the 
extent of private landownership in some of the major hydraulic 
civilizations. In this survey certain crucial data that were adduced in 
the discussion of our key criteria have to be mentioned again. But 
they now appear in a new context and, in a number of cases, they 
are supplemented by important additional information. In accord- 
ance with our previously established concepts we shall advance from 
simple to semicomplex and eventually to complex conditions of 
property and society. 


Hawaii: Ancient Hawaii certainly knew private possession of land. 
But it is doubtful whether there existed full landownership, since 
the "estates" even of the most powerful territorial "chiefs," the 
governors, "reverted to the king" after the holder's death and since 
"at the accession of a new king ... all the lands of an island" were 
reassigned. 32 

Inca Peru: As stated above, sinecure lands were held privately and 

CHAPTER 7, H 28l 

indefinitely, but the holders of such lands lacked the right to alienate 
them. Thus they were not owners but permanent occupiers. 

Sumer: At the close of the Sumerian period, genuine private land- 
ownership emerged. 33 However, the governments of the earlier 
temple cities seem to have exercised a strict control over the cul- 
tivable land. The records so far deciphered fail to reveal the exist- 
ence even of such private landed possessions as have been docu- 
mented for Inca society. 

Pharaonic Egypt: In addition to government land proper and to 
government-assigned land (temple land and office land), there was 
private land which could be alienated, 3 * but the king could cancel 
a holding at any time. Generally speaking, private landownership was 
more the exception than the rule. 35 


India: Numerous inscriptions document for the last southern phase 
of Hindu India what was already certain for the Buddhist and post- 
Buddhist periods, 36 namely that "most villages" were occupied by 
ryotwari 37 — that is, by peasants who were under the direct control 
of the state. This implies that private landownership can have 
existed only in a (not very large) minority of all villages. 

Mesopotamia: At the end of the Sumerian period and in Baby- 
lonian society private landownership is clearly apparent. Did it 
become the dominant form of land tenure? Evidence to this effect 
would lead us to class this period not as semicomplex but as com- 
plex. However, available data seem at best to indicate semicomplex 
patterns of property. At best. If state trade equaled or exceeded 
private trade during a large part of the Babylonian period, we would 
be faced with an advanced simple or an incipient semicomplex situa- 

For Late Sumer — the Third dynasty of Ur — the texts frequently 
mention private property in fields a well as in houses and gardens. 38 
But although the temples were no longer alone in leasing land, they 
are still most frequently mentioned in this respect. 1 For Babylonia, 
Meissner finds that the best and largest tracts of lands were in the 
hands of the government and the temples. "What still remained of 
the land was private property." 39 Schawe's analysis of the tenancy 
conditions of the period seem to confirm Meissner's view: in Baby- 
lonia land was rented out "primarily by the state and temple domains 
and then also by private individuals." *° 

i. Schneider, 1920: 58. Hackman's (1937: 21 ff.) numerous references to fields un- 
fortunately are often vague concerning their property position. 


Cuq stresses the specific position of land that was made private 
through royal gift. 41 At the same time he mentions among the 
features that led to differentiations in land tenure the emergence 
(or reemergence?) under the Kassites of communities which he char- 
acterizes as tribal and kinship-based, 42 but which he also compares 
with the Russian mir 43 — that is, with a purely administrative type 
of village community. The details of the Kassite rural units are 
still obscure, 44 but we know that they were interlinked with the 
governmental apparatus through certain of their leaders and that 
they regulated the landed possessions of their members in ways not 
too different from those of the Mexican calpulli and the Inca ayllus. 

During the last phase of Babylonian history the two types of 
government-controlled land still prevailed — if what the Persians 
found in Mesopotamia is indicative of the Neo-Babylonian condi- 
tions. In Persian Mesopotamia there were (1) state lands that in 
large part were assigne.- to individuals, (2) "large tracts of land" 
held by the temples, and (3) lands "held in fee simple by the in- 
dividuals." The first two categories were obviously very extensive: 
"With much of the land held by the state and the temples, the 
number of land transactions is not so large as that of other sales." 45 
Again we lack statistical data, but the above statement suggests that 
the process of "privatization" had gone further in mobile than in 
landed property. 

Persia: The Persians used the government-controlled land (and 
outside the Greek cities this was the bulk of all cultivated land) 
very much as had the Babylonians and Sumerians before them. They 
assigned it to members of the royal house and to friends of the king 
(obviously as sinecure land), to officials, resettled soldiers, and 
persons obligated to provide contingents for the army (obviously as 
office land). 46 Knowing the conditions under which office land was 
held in other Oriental despotisms, we have no reason to doubt that 
this land, like the regulated peasant land, was what Rostovtzeff takes 
it to be: state land. 

The Persian office land was not a feudal institution, nor did it 
inspire a feudal order among the Parthians. The big Parthian land- 
holders were no semi-autonomous fief-holders who spent most of 
their time attending to their personal affairs. Instead, and very much 
like their Persian predecessors, they were government officials. 3 

;'. Christensen, 1933: 307. According to Christensen, the Parthian government was 
"despotic" in form, at least as long as the Parthian monarchy was united (ibid.). 
Did the political order change when the monarchy disintegrated into several territorial 
kingdoms? This is, of course, possible but it is by no means certain. On the basis of 
comparable cases, it seems more likely that the smaller and later Parthian kingdoms 

CHAPTER 7, H 283 

Hellenistic Monarchies of the Near East: Private .landownership 
was confined essentially to the Greek cities, 47 which were few in 
Egypt but numerous in Western Asia. Outside these Greek enclaves, 
the land was controlled by the government and by government- 
attached temples. 

The Seleucid rulers established considerable private land through 
grant or sale 4S "on condition that the grantee joined his land to 
some city and made of it city-land"; 49 and, of course, they assigned 
office land to soldiers and probably also to civil functionaries. 50 

The kings of Pergamum do not seem to have reduced the royal 
land at all. "Like the Ptolemies, they must have gifted the (revocable) 
user of estates on King's land to officials." 51 

In Ptolemaic Egypt "private land originally meant house, gardens, 
and vineyard; even the house and garden of a Royal peasant were 
'private.' Greeks sometimes called it property, but it was, like every 
other Ptolemaic form, not property but user; apart from the Greek 
cities, the property or legal estate in any land in Egypt never left 
the king." 52 

It is in the light of this statement that we have to view the exist- 
ence of certain "private" grain land. Rostovtzeff suggests that this 
type of land had existed in Pharaonic Egypt; 53 and what we know 
of the earlier period confirms his assumption. However, we must 
remember first the instability that in Pharaonic times characterized 
landed property generally, and second the loose way in which the 
Ptolemaic (Greek) masters of Egypt employed the term "private." 

The "private" land, whose spread the Ptolemies encouraged, was 
"regulated emphyteutic" tenure 54 — that is, a lease of "deserted land" 
"for a long period (hundred years) or in perpetuity." The rights over 
this kind of property were "transferable by alienation or succession 
and enjoyed in a certain measure the same protection as owner- 
ship." 55 By developing emphyteutic tenure, the Ptolemies strength- 
ened the trend toward landownership. But until the Roman era, this 
trend does not seem to have gone beyond a relatively strong form of 
"landed possession." 56 

The Roman Interlude: Under the Romans private property 
emerged on a large scale. 57 The reasons for this extraordinary de- 
velopment — and for its limited success — are treated below in con- 
nection with the discussion of complex patterns of property. 

The Islamic Near East (the first centuries): The Arab conquerors 

were smaller Oriental despotisms, with certain leading families hereditarily holding 
the top-ranking positions in government and occupying very substantial tracts of 
office land. 


of Egypt and Syria perpetuated most of the Byzantine institutions, 58 
including patterns of land tenure. For obvious reasons many former 
holders of estates fled, 59 and those who stayed 60 lost the right to 
collect the taxes for the government. 61 Alongside them, prominent 
Arabs established themselves on deserted private estates and the 
old state domain. 62 These new holders bought and sold land, and 
they held their property, qati'a, 63 hereditarily. 64 But the qati'a repre- 
sented an emphyteutic form of possession; 65 and it is doubtful 
whether holders could enlarge them by freely buying peasant land. 
Their Byzantine predecessors had been forbidden by law to do 
so; 66 and the new Arab state was certainly as eager as the officials of 
Eastern Rome — and probably better able — to protect the regulated 
villages. Apparently, the qati'a possessions increased in extent, 67 but 
they remained in the hands of a limited group of leaders. The mass 
of the Arab tribesmen lived in military camps; 68 and it was only 
after several generations that the qati'a were spread into the villages. 69 

We need not follow here the step-by-step rise of a new system 
of landed property, whose beneficiaries were both tax collectors and 
holders of office land. 70 This system appears clearly and consistently 
in Mamluk society. 

Mamluk Society: At the beginning of Mamluk power virtually all 
the cultivable land of Egypt was divided into twenty-four units, 
which were either controlled directly by the sultan or assigned as 
office land. 71 Private land, mulk, was "almost absent." fc Its later 
growth was accomplished "mostly" by an intricate process which 
required a holder of office land to surrender part of it to the treasury 
before he purchased it from the government, either directly or 
through a middleman. 72 

But while mulk continued to increase until the end of the period, 
it remained only one of a number of types of land that an official 
(and usually a military official) might control. In addition to his 
office land (iqta') and to his mulk, he might possess pension land, 73 
and he might also be the manager of a waqf which he had founded 74 
and which in all likelihood would yield him and his family a steady 

Ottoman Turkey: The Turkish sultans demonstratively established 
the hegemony of state land by officially abolishing the bulk of 
privately owned land." 1 Some "landowners proper" seem to have 

k. Poliak, 1939: 36. Poliak assumes that private lands were, at the beginning of the 
Mamluk period, "numerous" in Syria. 

m. Gibb and Bowen, 1950: 236, 258, n. 4; cf. Poliak, 1939: 46. This refers essentially 
to cultivable land and pastures. The farmhouses and the land around it were always 
mulk; and vineyard and orchards were usually so considered (Gibb and Bowen, 
1950: 236). 

CHAPTER 7, H 285 

existed from the start; 7S and local "notables" (a'yans) acquired mulk, 
perhaps through conversion of office and other land. 76 But until the 
recent period of transition, most of the land was controlled by 
the government, which assigned part of it as office land or waqf 
and taxed the remainder through the agency of its tax farmers." 

The tax farmers had many prerogatives. In the non-Arab provinces 
they might transfer a vacant peasant farm ° to a resident of another 
village, but "only after offering it to the peasants of the village to 
which the land in question was attached." 77 In the Arab provinces 
their position, by the 18th century, approached that of holders of 
military office land. In Egypt they were given one-tenth of all village 
land under the name of wasiya. They could sell this wasiya land, but 
only to another tax farmer and only when they, at the same time, 
transferred to the buyer a corresponding amount of their jurisdic- 
tional domain. 78 In the Arab provinces the fellahs could alienate 
their land "to other fellahs." 79 Regarding the Arab territories, Gibb 
and Bowen expressly state that the person responsible for collecting 
the taxes "might not deprive a fellah of his land, except for non- 
payment of taxation." 80 Thus in both the non-Arab and Arab prov- 
inces the majority of all peasants were hereditary occupiers of assigned 
or regulated state land. 81 

The prerogatives of tax farmers and of holders of assigned lands 
present important problems; but all of them arise within the con- 
text of government-controlled land. Since land of this type comprised 
the bulk of all cultivated acreage, we feel justified in saying that the 
Islamic Near East, up to the 19th century, was characterized by a 
semicomplex pattern of Oriental property and society. 

Maya Society: The Maya system of land tenure is not clear. 8 - 
There probably was some individual landownership, 83 but most of 
the cultivable land seems to have been "common" (regulated) land. 84 

Pre-Conquest Mexico: Early sources agree that the bulk of all land 
in this area, as in Yucatan and Peru, was government controlled. 
The great majority of all peasants (and townspeople) lived in 
regulated communities (calpulli). 85 But there were also certain pri- 
vate lands, tierras proprias patrimoniales,** which were tilled by 
mayeques* 7 peasants attached to the soil. 

According to Zurita, private land had long been in existence. 88 
Did it originate through grant or sale? And how freely could those 
who held it dispose of it? Local officials were permitted to sell 

n. Gibb and Bowen, 1950: 237. The authors cite a statement according to which the 
"Sipahis" used to convert the state land they held into private property "in later 
times" (ibid.: 188, n. 6). Unfortunately the reference specifies neither the approximate 
date nor the extent of the development. 

0. A farm whose deceased owner was without heirs (Gibb and Bowen, 1950: 239). 


calpulli lands, if they were not burdened with obligations; and as 
stated above, the buyers of these tracts — which were then alienable — 
were either members of the ruling families or "some officials or 
merchants." 89 However, most of the calpulli land was burdened 
with serious and lasting obligations in that its yield was destined 
to support either the members of the calpulli themselves or officials 
of the local or central government, garrisons, or temples. 90 In con- 
sequence, the amount of land available for sale was probably small. 91 

It is not clear to what extent the tierras proprias patrimoniales 
originated from the sale of calpulli land. Some, or perhaps even 
many, of these private holdings may well have been grants made by 
the rulers to distinguished individuals. In contrast to the allodial 
estates of feudal Europe, the tierras proprias patrimoniales remained 
under the jurisdiction of the government; 92 and in contrast to the 
serfs of allodial or feudal estates the Mexican mayeques served the 
government "in time of war or need." 93 This formula is compre- 
hensive. In Aztec Mexico, as in other hydraulic societies, the govern- 
ment determined one-sidedly what kind of services it needed. 

Not being office lands, the private holdings were not kept intact 
by the will of the government. And not being allodial or feudal 
estates, they were not entailed by the will of the owner: "no son de 
mayorazgo." 9 ' 4 In fact, the private lands of ancient Mexico were as 
similar to the sinecure lands of other Oriental societies as they 
were dissimilar to the strong landed property of feudal and post- 
feudal Europe. In all probability they represented a smaller per- 
centage of all cultivated land than did private lands in Babylonia 
or in early Islamic society. According to one estimate, the private 
holdings in ancient Mexico amounted to little. 95 According to an- 
other, they may have comprised somewhat more than 10 per cent of 
the total cultivated area.* 


Oriental societies in which there was less private land than 
government-controlled land are many. Private land was insignificant 
in the higher civilizations of South and Meso-America when they 
were overrun by the Spaniards. It remained a secondary feature in 
India, Sumer, Babylonia, Persia, the Hellenistic monarchies of the 
Near East, and Islamic society. In the early phases of state-centered 
Chinese society it appears to have been as unimportant as it was in 

p. This figure was suggested in a memorandum on land tenure in pre-Conquest 
Mexico prepared for the present inquiry by Dr. Paul Kirchhoff. 

CHAPTER 7, H 287 

pre-Conquest America; and when China, under the impact of Inner 
Asian forces, temporarily discarded the free forms of landed property 
that had prevailed at the end of the Chou period and throughout 
the imperial dynasties of Ch'in and Han, regulated patterns of land 
tenure prevailed again. 

Thus, our survey confirms what we tentatively suggested at the 
start of our discussion of hydraulic land tenure. Prior to the recent 
period of institutional disintegration and transition, private land 
may have prevailed in the Near East under Roman rule; it certainly 
prevailed in China from the later part of the first pre-Christian 
millennium to the 5th century a.d. and, after an interlude of almost 
three centuries, again and until our time. 

The Roman Near East: Did such classically hydraulic countries as 
Egypt, under Roman rule, actually develop complex patterns of 
property? The conquerors did indeed establish private landed prop- 
erty within the terms of Roman provincial law; 96 and in Byzantine 
Egypt prior to the Arab conquest large estates were certainly held 
by the "powerful ones," the dynatoi. But how widespread was land- 
ownership at the beginning of the Roman empire? And to what 
extent did it prevail during the 5th and 6th centuries? 

Under Roman influence private land was created through grants, 97 
through transfer of cleruchic land (military office land), 98 and through 
the sale and grant of other government land. 99 This was a far cry 
from Hellenistic conditions; but even scholars who emphasize the 
qualitative differences 10 ° are usually careful also to indicate the 
quantitative limitations. The greater part of the former cleruchic 
land was taken back by the government immediately after the con- 
quest; 101 and out of the private estates that temporarily came into 
being as the result of grants or sales "the majority" soon again 
became imperial property. 102 Thus "the best land continued for the 
most part to form the royal domain and to bear the name royal 
land." q And since, in the main, it was the larger estates that were 
confiscated, private land seems essentially to have been held by small 
owners. This is particularly true for Egypt and Asia Minor. A 
greater incidence of large estates is suggested for Syria and Pales- 
tine. 103 

The existence of private landownership is said to have reached a 

q. Bell, 1948: 73. On the basis of several decades of additional research Bell con- 
firms what Mommsen had cautiously noted in 1885, namely that the imperial domain 
constituted "a considerable part of the entire area in Roman as in earlier times" 
(Mommsen, 1921: 573). Johnson and West (1949: 22) refer to "the retention of the 
great bulk of arable land as the property of the [Roman] crown"; and Johnson (1951: 
92) calls "the amount of privately owned land in Roman times . . . slight." 


second peak on the eve of the Arab conquest, especially in Byzantine 
Egypt. What actually were the conditions of land tenure in Egypt 
during this period? The peasants, who because of extreme fiscal 
pressure had become increasingly reluctant to farm — not a few ran 
away from their villages — became the targets of elaborate "reform" 
measures. Government control in the form of compulsory permanent 
tenancy (epibole) became more and more strict. 104 Increasingly the 
peasants were permanent holders of land which they were forbidden 
to leave. As coloni, they were attached to the land which, from 
then on and within the confines of a rigidly regulated village com- 
munity, became their "private" possession. 105 The continuing fiscal 
burdens caused many villages to look to "powerful" protectors, 
primarily members of the governing class, and to the church. 100 
These individuals, who were designated patroni until 415, 107 did not 
exert authority everywhere — many villages remained directly sub- 
ordinated to the fiscus and the imperial administration. 108 Nor did 
they integrate "their" peasants into a typical and large-scale manorial 
economy, 109 although for lack of a better term their holdings are 
usually referred to as "estates." 

The edict of 415, which acknowledged the position of large land- 
holders, also reaffirmed the government's claim on the fiscal and corvee 
services that the landholding coloni had previously fulfilled. 110 The 
holders of the new estates were delegated to collect taxes for the 
government from their coloni. But although this function endowed 
the new landlords with great power, 111 the state upheld its fiscal 
rights without compromise: "the rate of taxation was the same for 
all." 112 Thus with regard to the most crucial fiscal aspect the estate 
holders were not privileged: "that their tax rate was less than others, 
there is no evidence whatsoever." r 

Under Justinian (to be precise: in 538) the Byzantine government 

r. Johnson and West, 1949: 240. In the 2d and 3d centuries the tax collectors seem 
to have been in the main municipal groups or individual businessmen who had their 
fiscal duties imposed upon them as a "liturgy." The government used these liturgical 
obligations to destroy the economic strength of propery-based groups (Wallace, 1938: 
347 ff.); and it transferred the fiscal tasks to the bureaucratic landlords, who, being 
politically better connected, succeeded where the private entrepreneurs had failed. But 
these landholders were in no sense feudal lords, who could appropriate the bulk of 
the peasant surplus they collected, From the 4th to the 6th century the Byzantine 
collectors were generally allowed commissions of some 2 per cent on the collection 
of wheat, 2% per cent on barley, and 5 per cent on wine and pork (Johnson and 
West, 1949: 328, cf. 290). Whether these rates were valid for Egypt we do not know 
(ibid.); but we do know that the Egyptian tax collector was entitled to a fee of one- 
eighth to one-twelfth of the money tax he raised (ibid.: 268, 284), that is, to a com- 
mission of 8 to 15 per cent. By manipulation he could raise his share to from 10 to 
20 per cent of the money tax (ibid.: 268, 284 ff.). 

CHAPTER 7, H 289 

expected a tax revenue from Egypt that was larger than that men- 
tioned for the time of Augustus. 113 This fact involves a number of 
questions which have not as yet been solved. 114 For our purpose, 
however, it is enough to know that the Byzantine government was 
able to tax the Egyptian peasants as comprehensively and success- 
fully as did the Romans under their powerful first emperor. 

To be sure, there were in Egypt at the close of the Byzantine 
period large units of private landed property: estates. These estates 
arose under a bureaucratic government; they were held mainly by 
bureaucratic landlords; and they were organized in a conspicuously 
bureaucratic way. 8 

All this we know. We do not know, however, "whether these estates 
of Egypt were privately owned or were leaseholds from imperial and 
ecclesiastical properties, or even from small farmers." 115 We do not 
know either whether these estates, prior to the Arab conquest, com- 
prised more than one-half of the cultivable land. The law forbade 
the estate holders to purchase peasant land at will, and according to 
Johnson, 116 "there is no evidence" that this legislation "was ever 
a dead letter." The landlord's proprietary position, even if it had 
the character of ownership, was legally limited. The freedom of the 
villagers, it need scarcely be repeated, was even more severely re- 

The historical data known today suggest that in such Near 
Eastern countries as Egypt private landownership did not prevail 
at the beginning of the Roman period and they give little reason to 
assume that this type of ownership spread later in such a way as to 
even temporarily establish complex patterns of property and society. 

China: Authentic historical records state that in the 4th century 
B.C. in the state of Ch'in the traditional regulated field system was 
abolished and that from then on, land could be bought and sold 
freely. 117 The records dealing with the imperial dynasties of Ch'in 
and Han imply that after the unification of China private land- 
ownership prevailed generally. 118 When, in the first century B.C., 

s. Bell finds that in contrast to feudal lordship in the West, which "was a replica 
in little of the kingdom to which it belonged," the estate of Byzantine Egypt "repro- 
duced in little the bureaucratic empire of which it formed a part; its organization and 
its hierarchy of officials were modelled on the Imperial bureaucracy. Indeed it is 
sometimes impossible, in dealing with a papyrus document of this period, to be 
certain whether the persons whose titles are mentioned in it were Imperial officials 
or the servants of some great family" (Bell, 1948: 123 ff.). This overlapping of titles, 
far from being accidental, reflects an overlapping in positions. The proprietors of 
these estates were for the most part, if not exclusively, officiating or nonofficiating 
members of the governing class, who even in their capacity as landlords functioned 
as semi-officials: tax collectors and leaders of the hydraulic and nonhydraulic corvees. 


merchants accumulated substantial mobile and immobile property, 
the government took strong fiscal measures to reduce their wealth, 
and an edict in 1 19 B.C. forbade them to own land; 119 but this edict 
did not interfere with land transfers between other classes, and 
even in the case of the merchants it seems to have been maintained 
only temporarily. 

Unfortunately the historical sources leave important aspects of 
the agrarian development unexplained; and this is true both for the 
first period of complex property relations and for the subsequent 
regulated agrarian order that was instituted in the 5th century a.d. 
and endured until the middle of the 8th century. However, the 
information at hand is sufficient to illuminate at least the main 
trends in these periods. 120 During the last millennium, dynasties of 
conquest reserved lands for their tribal supporters and for some 
Chinese who had joined their conquering armies; but for the bulk 
of their Chinese subjects they upheld private landownership. It has 
been estimated that during the last phase of the Ch'ing (Manchu) 
dynasty the combined bannerland of the Manchu, Mongol, and 
Chinese bannermen amounted to some 4 per cent and privately 
owned land to almost 93 per cent.' 

Although prior to this phase nongovernment land may at times 
have amounted to no more than one-half of all land, 121 and although 
a variety of legal clauses gave the right of preemption (primarily) 
to relatives, 122 it seems evident that China went further than any 
other major Oriental civilization in maintaining private ownership 
in land. 

The reasons for this extraordinary development are by no means 
clear. But certain facts are suggestive. In China the critical changes 
occurred after the middle of the first millennium B.C., when several 
important culture elements appeared simultaneously: plowing with 
oxen, the use of iron, and the art of horseback riding. We hesitate 
to dismiss this coincidence as inconsequential. None of these elements 
emerged in the hydraulic areas of pre-Conquest America; and in the 
Near East and India they emerged separately in the course of a 
drawn-out development. In both areas plowing with labor animals 
was known from the dawn of written history, whereas the use of 
iron spread later, and the art of horseback riding later still. Could 

t. Buck, 1937: 193. The estimate used by Dr. Buck puts privately owned land at 
92.7 per cent, land assigned to Manchu nobles together with some "crown land" at 
3.2 per cent, "state land" (land set aside for the maintenance of schools, religious 
purposes [state cult]) at 4.1 per cent. These data are approximate. They do not make 
allowance for private ancestral and temple land, which, according to the same source, 
amounted to less than .05 per cent. 

CHAPTER 7, H 291 

it be that the simultaneous rise of new techniques of agricultural 
production and of military coercion and fast communication (and 
the assurance the two last gave to the maintenance of government 
control) encouraged the masters of Chinese society to experiment 
confidently with extremely free forms of landed property? Whatever 
the reason for the fateful step may have been, once taken it was 
found to be politically workable and agronomically and fiscally 

The Chinese development — which requires further investigation 
— is remarkable not only for its success but also for its geographical 
limitations. It seems to have affected certain southwestern neighbors, 
especially Siam. But many cultural contacts with more remote Asian 
countries notwithstanding, the Chinese system of private landowner- 
ship remained essentially confined to the area of its origin. 

3. How Free Is Private Landed Property 
in Hydraulic Society? 

Thus private landownership was present in many hydraulic civiliza- 
tions; but except for a brief and recent period of transition, the 
combined private lands were less extensive than the combined public 
lands. More. Even where private landownership did prevail, it in- 
variably was prevented from achieving the kind of freedom which is 
possible in a multicentered nonhydraulic society. 

a. Despotically Imposed versus Democratically Established 
Restrictions of Private Property 

To be sure, in no society does an owner absolutely dispose over his 
property. Even under conditions of strong property the owner of 
bricks, who may sell or store them or use them in building his house, 
may not throw them at his neighbor. The early Roman emphasis 
on the proprietor's sovereign position, although meaningful fiscally, 
is not valid societally. 

Even fiscally the holder of strong property is not necessarily 
without burdens. In most free commonwealths some public func- 
tionaries have to be supported, and when this is the case, the citizens 
may have to draw upon their property to satisfy this need. Contribu- 
tions from private property for the maintenance of the government 
will be used only for proven essentials when the property-based 
forces of society can keep government in a serving position. Such 
contributions will increase, and be spent more freely, when an im- 
perfectly controlled government partially determines its own budget. 
They will be determined one-sidedly and with primary concern for 


the interests of those in power when a state stronger than society 
prevents the representatives of property from protecting their inter- 
ests. It is under conditions of the first type that we find strong, 
though never absolute, property. And it is under conditions of the 
third type that property is weak. In hydraulic society immobile 
property, like mobile property, remains weak even where private 
landownership quantitatively outweighs public land tenure. 

b. Restrictions Imposed upon the Freedom to Enjoy, to 
Use, to Transfer, and to Organize 

Oriental despotism one-sidedly restricts the landowner's freedom 
to enjoy the fruits of his property, to decide on its use, to will it 
freely (through testament), and to protect it by means of political 

The agrodespotic government demands payments from all land- 
holders, either for its own use or for the use of especially privileged 
persons or institutions (temples, mosques, churches); and it deter- 
mines the land tax one-sidedly, according to its own (the rulers') 
rationality standard. Tenancy may stratify the proprietary sector; and 
the changing strength of local and central authorities may alter the 
distribution of state revenues within the bureaucratic order. But 
neither condition affects the fundamental arrangement that compels 
owners and/or possessors of land generally to surrender a substantial 
part of their revenue to the representatives of the state. 

Directly, this arrangement aims at the fruits of operational landed 
property. Indirectly, it also influences (and limits) the use to which 
a given piece of land may be put. The government bases its fiscal 
demands on the expectation that the peasant occupiers (or owners) 
will grow a crop capable of yielding a certain return. This demand 
forces the cultivator to grow the standard crop or an acceptable sub- 
stitute. Occasionally, and particularly in regulated agrarian orders, 
the government may expressly prescribe that certain plants or trees 
(rice, corn, olives, hemp, cotton, or mulberry trees) be cultivated; 
and in these cases the proprietor's freedom to determine how his land 
should be used is nil. Frequently, however, the government is con- 
tent to prescribe how much should be paid over to it. In both cases 
the result is a crude type of planned economy, which substantially 
limits the cultivator's freedom of choice and action. 

Restrictions on the freedom to will property and to organize for its 
protection have been discussed in an earlier chapter. Hydraulic laws 
of inheritance fragment privately owned land. The landowner's in- 
ability to strengthen his proprietary position through independent 

CHAPTER 7, I 2 Q2 

national and politically effective organizations is as apparent in com- 
plex as in semicomplex or simple hydraulic societies. 

This does not mean to say that the predominance of private land- 
ownership and the spread of landlordism in such civilizations as tra- 
ditional China were societally irrelevant. They were not. But the 
spread of landlordism, which significantly modified the relations be- 
tween the officiating and nonofficiating (gentry) segments of the rul- 
ing class, did not result in the consolidation of landed property or in 
independent organizations of landed proprietors. From the fiscal, le- 
gal, and political points of view private landownership was as weak 
at the final collapse of traditional Chinese society as it had been at 
its birth. 


1. The Perpetuation of Hydraulic Society 
Depends on the Government's Maintenance 
of its Property Relations 

On the basis of these facts certain general conclusions seem justified. 
First of all, hydraulic society, like other institutional conformations, 
knows private property. Human existence over any considerable 
period of time is impossible without the public recognition and 
standardization of relations between persons and things or services. 
Even the convict possesses his clothes while he wears them; and 
many slaves possess not only their clothes but certain other articles as 
well. A serf possesses a great variety of things in addition to his land. 

In most cases possession — and, of course, ownership — are recog- 
nized by custom. Where written laws exist, important forms of 
property may be recognized and regulated by special statutes. 

This is true for all societies, including those ruled by despotic 
regimes. The most elementary considerations of rationality require 
that even those who make — and change-— laws one-sidedly and 
despotically should emphasize their validity by not abrogating them 
unnecessarily. A ruler's rationality coefficient is the higher, the more 
strictly he himself observes the regulations which he has imposed 
upon his subjects. This also includes regulations concerned with 
private property. 

The Oriental despot may buy and sell land. 1 He may have private 
artisans producing goods for him and at times he may pay them 
generously. And he may also buy directly from merchants. In all 
these cases he may — though he need not — set a low price. In Mus- 
covite Russia this seems to have been the rule, 2 and in classical Hindu 


India merchants had to accept whatever figure the king's appraiser 
deemed appropriate. 3 But the fact that the ruler and his officials 
paid for certain goods and services does not negate the despotic 
character of the regime. It only shows that by and large the despotic 
regime proceeds on the basis of the legal and proprietary regulations 
it has established. 

What is true for Oriental despotism is no less true for the modern 
industrial apparatus state. Superficial observation may be satisfied 
with the presence of laws that deal with property. But no realistic 
analyst will call the Hitler government democratic because it dealt 
with Jewish property in accordance with the Nuremberg laws. Nor 
will he deny the absolutist character of the early Soviet state because 
it bought grain at a government-fixed price from individually pro- 
ducing peasants. 

2. The Growing Complexity of Property and 
the Growing Complexity of Society 

In addition to being an essential feature of hydraulic society, hy- 
draulic property is also characterized by a variety of forms. Con- 
siderable private property and enterprise may appear in industry 
and commerce; and private ownership may spread, and even prevail, 
in agriculture. The representatives of semicomplex and complex 
patterns of property maintain relations to one another and with the 
state that differ substantially from those maintained by representa- 
tives of simple patterns of property. This fact enables us to dis- 
tinguish, on the basis of different patterns of property, different 
subtypes of the over-all societal order. 

3. Small Property Offers a Considerable Economic 
Incentive, but No Political Power 

a. Incentives Inherent in Private Possession and Ownership 

The technical advantages accruing from devices that can be em- 
ployed only by large teams may equal or outweigh what is achieved 
by individual effort or by the labor of a few kinsmen working to- 
gether. But when the technical advantages are insignificant or lack- 
ing, the incentives for individual action tend to become more 

Individual action need not be based on ownership. The occupier 
of a piece of land may only be its possessor, but in premachine days 
and under comparable technical conditions he is likely to outproduce 
a member of a team who is working for hire. Throughout the 

CHAPTER 7, I 295 

hydraulic world we therefore find the peasants tilling their land in- 
dividually rather than collectively; and where labor animals increased 
the advantages of individual cultivation, small-scale peasant work 
also replaced the only relevant system of collective agriculture, the 
public field system. In handicraft and commerce private enterprise 
is generally based on private ownership. In agriculture private 
possession is usually sufficient to make the peasant proceed with 
great care. Tenancy, like peasant ownership, has created a horticulture- 
like intensity of farming. 

To be sure, the tenant's desire to own his land is enormously 
strong. Even under the most frustrating fiscal conditions most peasant 
owners cling to their fields in the hope that the irrational tax 
pressure will be lightened before they are forced to abandon their 

Property-based private handicraft created many of the beautiful 
objects (textiles, wood-, leather-, and metal-work) that delight the 
student of hydraulic civilization; and the hydraulic peasants who 
individually tilled their fields surpassed in skill and productivity the 
serfs of Medieval Europe. This was so even when these peasants were 
only the hereditary occupiers of regulated land; and it was even 
more so when they were tenants or private landowners. Indeed it is 
not at all unlikely that the exceptional intensity of agriculture in 
traditional China derived from the fact that private peasant land- 
ownership was more widespread there than *in any other major 
hydraulic civilization. 

b. The Beggars' Property 

Small private property, both possessed and owned, was conspicuous 
in hydraulic societies of the semicomplex type. It became much more 
so, and particularly in the agrarian sphere, in complex Oriental 
societies. Did it in either case become an important political force? 
From the standpoint of a multicentered property-based society, 
the question is entirely reasonable. Small proprietors (artisans and 
peasants) played an ever increasing political role in classical Greece. 

a. The feudal landholders of Japan did not engage in large-scale manorial farming 
as did their European peers; and the Japanese peasants cultivated their land in- 
dividually and under conditions which resembled tenancy rather than serfdom. On the 
basis of a highly refined irrigation economy, they too engaged in a semihorticultural 
type of farming. This cannot be explained entirely by geographical proximity. The 
Japanese did not adopt the semimanagerial bureaucratic absolutism of China; nor did 
they adopt their system of private landownership from their continental neighbors. 
But within a feudal framework of power and social relations, the Japanese nobles 
gave their peasants as great a proprietary incentive as the over-all pattern of their 
society permitted. 


Independent artisans were prominent in many guild cities of Medi- 
eval Europe; and, together with the peasants, they constituted a 
significant element in the democratic governments of Switzerland. 
In a number of the predominantly agrarian states of the United 
States which are not given over to giant farms and large-scale pro- 
duction the farmers' vote is a decisive factor. Although today the 
farmers account for less than 1 5 per cent of America's manpower, they 
are better organized than ever, and they continue to be a substantial 
political force, both regionally and nationally. 

There is no need here to stress the potential political importance 
of labor — a group whose essential economic asset is the capacity to 
work. Free labor became a political force in ancient Greece in a num- 
ber of city states during the final phase of their independence. Or- 
ganized both professionally and politically, the representatives of this 
form of individual property have in some industrial countries, such as 
Australia, Sweden, and England, assumed political leadership even in 
matters of the national economy; and in many others, including the 
United States, their political position has improved rapidly. 

Small property and labor played no comparable role in the hy- 
draulic world. With regard to labor the issue is simple. Personally 
free, hired laborers have existed in many hydraulic civilizations. 4 
Unskilled workers were for the most part unorganized. Skilled 
workers were frequently organized in local and separate professional 
units,. But even when they were not under strict government super- 
vision, they constituted only a politically irrelevant form of self- 
government, a Beggars' Democracy. 

And the peasant proprietors? Whether they possessed or owned 
their land, they remained the representatives of a fragmented type 
of property and enterprise. At best they were permitted to handle 
their essentially local affairs within the rural version of a Beggars' 
Democracy, the village community. 

From the standpoint of the absolutist bureaucracy, the property 
of both artisans and peasants was Beggars' Property, property that 
was economically fragmented and politically impotent. 6 

b. Did the peasants constitute an economic and political threat during the first period 
of the Soviet regime? Long before 1917 Lenin stressed the danger of any private 
property (peasant land included) for a socialist regime (cf. Lenin, S, IX: 66-7, 213-14, 
and passim); and he did not change his opinion after his party established its 
dictatorial power with the support of the peasants who had been "given" land (ibid., 
XXVII: 303 ff.; XXXI: 483 ff.). He insisted that property transforms men into "wild 
beasts" (ibid., XXX: 418); and he called the petty bourgeois and small peasants 
potential breeders of capitalism and thus an inherent danger to the Soviet state (ibid., 
XXVII: 303 ff.; XXXI: 483). In 1918, and again in 1921, Lenin viewed these petty 

chapter 7, i 297 

4. Private Commercial Property Politically Inconse- 
quential Even When Permitted to Become Large 

Under certain conditions, the representatives of Oriental despotism 
found it economically advisable to have the bulk of all trade handled 
by private businessmen. When this was the case, some merchants 
grew fabulously rich, and a few enjoyed distinguished social posi- 

We do not exclude the possibility that big merchants as a group 
could have participated in the running of despotic governments; but 
the evidence at hand fails to document this development as a signifi- 
cant feature in any of the major representatives of semicomplex or 
complex hydraulic societies. In Babylonia, in Buddhist India, in pre- 
Conquest Meso-America, in the Islamic Near East, and in imperial 
China big merchant property, even when involved in large-scale 
operations, remained politically inconsequential. 

5. Problems of Wealth within the 
Governing Class 

Property problems of a very different kind arise within the govern- 
ing class. In simple hydraulic societies almost the entire national 
surplus is appropriated by the ruler and his serving men. And even 
when intermediary groups, such as the merchants, are permitted to 
derive considerable profit from their transactions, the governing class 
continues to monopolize the greater part of the country's wealth. 
The members of the court and the officials may receive their share 
of this wealth either as revenue from assigned (office or sinecure) lands 
or as salary (in kind or cash). In both cases the income is based on 
the government's power to control land and to tax people. And in 
both cases it becomes private (bureaucratic) property. Its recipients 
may use all of it for consumption; or they may set some part of it 
aside as savings or for investments. Both types of use involve the 
problem of bureaucratic hedonism; the second raises in addition the 
problem of bureaucratic landlordism and capitalism. 

bourgeois forces as his regime's "principal enemy" (ibid., XXXII: 339). On the eve 
of the First Five- Year Plan, Stalin repeated Lenin's formula that the small producers 
are "the last capitalist class" (ibid., XXXII: 460). He insisted that this class "will 
breed capitalists in its ranks, and cannot help breeding them, constantly and con- 
tinuously" (Stalin, 1942: 102). There is no evidence that either Lenin or Stalin based 
his views on a serious study of the political position of small peasant landholders under 
absolutist state power. Stalin's pseudoscientific accusations of 1928 only served to pre- 
pare the Soviet bureaucracy and the Soviet people for the total liquidation of private 
peasant property. 


a. Bureaucratic Hedonism 

Bureaucratic hedonism can be defined as the enjoyment of 
wealth without provoking the envy of high officials or the crushing 
wrath of the despot. 6 Such hedonism may be complicated by op- 
portunities for saving and investing. While the members of the 
governing class are generally eager to enjoy their property as long 
as the enjoying is good, they express this desire differently under 
different circumstances. But the wish to consume pleasantly and to 
live well prevails everywhere, even in those complex hydraulic 
societies in which the possibility of owning land encourages economy 
and thrift. Often and particularly in the case of very highly placed 
and continually endangered officials, such as viziers, chancellors, or 
"prime ministers," the bureaucratic joie de vivre is spectacularly 
manifested. 6 

b. Bureaucratic Landlordism and Capitalism 

Even the most luxury-loving functionary usually tries to save part 
of his income. After all, he may not be in office forever; his family 
will always have to eat; and his children will have to be trained 
for the most desirable of all goals: a government career. Thus the 
thoughtful official buries precious metals and jewels in the ground. 
Or better still he converts some of his passive private property into 
active property. He buys land for rental and/or he uses his funds 
profitably as a government contractor (especially as a tax collector), 
or as a money lender, or as a partner in private commercial enter- 
prise. On the basis of his bureaucratic property he becomes a 
bureaucratic landlord and/or a bureaucratic capitalist. 

Of course, there are others who are landlords also. Wherever land 
can be freely alienated, small proprietors are eager to purchase it. 7 
And there may also be nonbureaucratic capitalists. But since the 
uniquely powerful state apparatus surpasses all other forces of hy- 
draulic society in acquiring agricultural and nonagricultural revenue, 
officials figure prominently as tax farmers and, wherever land can 
be bought, as landowners. 

In imperial China predominance of private landownership en- 
abled the officials to invest a considerable part of their income in 
land. A recent analysis of officialdom and bureaucratic gentry in 
lgth-century China suggests that at the close of the Ch'ing dynasty 
present and former officials, holders of official titles, and holders of 

c. The pleasures of bureaucratic consumption in late imperial China were depicted 
in great detail in such novels as the Dream of the Red Chamber. 

CHAPTER 7, I 299 

high examination degrees together may have received land rents 
amounting to 165 million taels annually and about 81.5 million taels 
from entrepreneurial activities. At the same time all lower degree- 
holders together received about 55 millions from the first source and 
40 millions from the second. 4 These figures indicate that by far the 
greater part of the rent income of the official-literati went to the 
upper echelon of this group, which was predominantly bureaucratic 
(present and former officials and quasi-officials outweighed the holders 
of high degrees by three to one). 6 

Under Oriental as well as under Occidental despotism, landlord- 
ism and officialdom overlap. But the seemingly similar configura- 
tions differ profoundly- in their institutional substance. The bureau- 
cratic landlords of Oriental society derived their political power 
essentially from the absolutist government, of which they themselves 
or their officiating relatives formed an active part. It was only as 
officials that the members of the agrobureaucratic gentry were politi- 
cally organized. The noble landlords of postfeudal Europe or Japan 
did not necessarily hold government office. And they did not need 
government salaries to periodically restore their landed property, 
since their estates were kept intact by primogeniture and entail. 

The land of the bureaucratic (Oriental) gentry might facilitate a 
government career for certain of its members and thus give renewed 
access to power; but essentially this land was revenue property. 
Conversely, the land of the feudal (Occidental) gentry involved the 
perpetuation of organized political power, independent of, and at 
times openly conflicting with, state power. In a way unparalleled 
by hydraulic property (bureaucratic and other) and in addition to 
being revenue land, feudal property was conspicuously and signifi- 
cantly power property. 

d. These and many other illuminating data have been taken from a comprehensive 
study of the Chinese "gentry" in the 19th century by Dr. Chang Chung-li, University 
of Washington, Seattle, who has generously permitted their inclusion in the present 
study. The officials and degree-holders are classed together, because during the later 
part of imperial China they constituted a status group, shen-shih (see below, Chap. 
8). Their entrepreneurial income stemmed mainly from investments in native banks, 
pawnshops, and the salt trade (Chang, GI, Pt. II). Dr. Chang's study shows that the 
shen-shih — "a privileged group with managerial abilities and functions" — received 
"from 'government services,' 'professional services' and 'gentry services' together" an 
income larger than that from rent or mercantile activities (letter from Dr. Chang, 
March ao, 1954). 

e. Prior to the Taiping Rebellion, the "officials, officers and holders of official titles" 
together constituted 67 per cent of the upper group; after the Taiping period, the 
figure rose to 75 per cent (Chang, CG, Pt. II). 

goo patterns of proprietary complexity 

6. Conclusions Leading to New Questions 

a. Hydraulic Property: Revenue Property versus Power 

Whether hydraulic property is large or small or whether or not 
it belongs to a member of the governing class, it provides* material 
advantages. But it does not enable its holders to control state power 
through property-based organization and action. In all cases, it is 
not power property but revenue property. 

b. The Importance — and Limitation — of Private Property in 
Determining Class Differentiations within Hydraulic Society 

This does not mean to deny the importance of property in establish- 
ing social (class) differentiations. The emergence of property-based 
handicraft and commerce and the spread of private landownership 
involve the emergence of new social elements, groups, and classes. 
Thus it is not only legitimate but necessary to show in which ways 
patterns, of social differentiation correlate with patterns of private 

However, it is quickly apparent that in hydraulic society the 
problem of social differentiation involves more than the question 
of the presence or extent of private property. Once established, 
bureaucratic wealth is private property, but it is rooted in, and 
derives from, government property, and its intrabureaucratic dis- 
tribution is based on political conditions that cannot be explained 
in terms of private property. 


/ lasses in hydraulic society 


Modern institutional analysis emerged in a society that was deci- 
sively shaped by conditions o£ property. Consequently, the pioneers 
in the modern sociology of class saw the major segments ("orders") x 
of society as determined essentially by major types of private property 
and by corresponding types of revenue. According to Adam Smith, 
"the whole annual produce of the land and labour of every country 
. . . naturally divides itself . . . into three parts; the rent of land, 
the wages of labour, and the profits of stock; and constitutes a 
revenue to three different orders of people; to those who live by 
rent, to those who live by wages, and to those who live by profit. 
These are the three great, original and constituent orders of every 
civilized society, from whose revenue that of every other order is 
ultimately derived." 2 The representatives of government are sup- 
ported to an extent from "public stock and public lands"; but the 
greater part of their expenses is met by the three major orders which 
render some of their revenue to the state in the form of taxes. 3 

According to this view, the representatives of government con- 
stitute not a major order of society but a secondary and derivative 
one. And whenever conflicts concerning property arise, civil govern- 
ment becomes a weapon of the propertied classes against the eco- 
nomically under-privileged groups. To quote Smith again: "Civil 
government, so far as it is instituted for the security of property, 
is in reality instituted for the defence of the rich against the poor, 
or of those who have some property against those who have none 
at all." a 

This statement, which was written in a period of unbridled 

a. Smith, 1937: 674. Smith supplements this statement by a citation from his "Lec- 
tures": "Till there be property there can be no government, the very end of which 
is to secure wealth and to defend the rich from the poor." He adds a reference to 
Locke, Civil Government, sec. 94: "Government has no other end but the preservation 
of property." 



proprietary privilege, presents a crude economic interpretation of 
the state. It makes no allowance for power as an independent deter- 
minant of class or for the socio-economic prominence of the state in 
the hydraulic civilizations with which Smith was familiar. 6 Smith's 
successors defined the peculiarity of Asiatic society more clearly; 
but they, too, treated "Asia" as a residual category in a socio-eco- 
nomic system which considered private property and the revenue 
derived from it the decisive factors in the formation of class. 

Despite its obvious deficiencies, the proprietary concept of class 
greatly stimulated the social sciences up to the beginning of the 20th 
century. Without doubt, this concept is essential for the under- 
standing of societies in which strong independent private property 
prevails; and it remains important also for the understanding of 
certain secondary aspects of power-based societies. But it is insuffi- 
cient when it is unqualifiedly applied to formations of the first type. 
And it is altogether inadequate when it is used as the essential means 
for explaining formations of the second type. 

The growth of big government in many modern industrial coun- 
tries and the rise of totalitarian states in the USSR and Germany 
enable us to recognize state power as a prominent determinant of 
class structure, both in our time and in the past. They also enable 
us to recognize more clearly than before the importance of power 
in the establishment of the ruling class in hydraulic society. 


1. The Key Criterion: Relation to 
the State Apparatus 

The pioneers of a property-based sociology of class viewed the 
Asiatic state as a gigantic landowner. In most hydraulic societies the 
bulk of all cultivated land is indeed regulated; and although the 
state's proprietary right over the regulated fields is hidden behind 
the facade of a seemingly self-governing village community, it 
operates negatively when the government prevents outsiders from 
purchasing these fields, and positively when the government assigns 
or sells land (or villages) at will. However, the classical formula is 
definitely unsatisfactory in at least one respect: It overlooks irriga- 
tion water, which in hydraulic societies is a major agent of produc- 

b. Smith, 1937: 789 ff. On a number of occasions Smith tries to remove the incon- 
sistency by limiting his scheme to "civilized" societies. But he makes no effort to 
establish a concept of class that adequately mirrors the specific position of the state and 
its representatives in either the Eastern or the Western world. 

CHAPTER 8, B 303 

Does the despotic state "own" the great accumulations of water? 
This has been claimed in many but not all hydraulic civilizations. 
I prefer to view the state as controlling rather than owning the 
country's "big" water. 

The same approach may also be taken with regard to land. Some 
hydraulic states, such as imperial China, tolerated the predominance 
of privately owned land over a long period of time, and in this case 
the state restricted the owner's proprietary position by means of 
heavy taxation, directives as to what crops should be grown, and a 
fragmenting law of inheritance. Thus the hydraulic state, which 
frequently owned the bulk of all cultivable land, generally kept 
landed property weak. Its position is again best viewed as one of 

In hydraulic society the first major division into an order of 
superior and privileged persons and an order of inferior and under- 
privileged persons occurs simultaneously with the rise of an in- 
ordinately strong state apparatus. The masters and beneficiaries of 
this state, the rulers, constitute a class different from, and superior 
to, the mass of the commoners — those who, although personally 
free, do not share the privileges of power. The men of the apparatus 
state are a ruling class in the most unequivocal sense of the term; 
and the rest of the population constitutes the second major class, 
the ruled. 

Within the ruling class different individuals and groups differ 
greatly in their ability to make decisions and handle personnel. In 
the civil administration, as in the army, major directives originate 
at the top level. But, again, as in the army, minor decisions are made 
by men in the middle brackets. And decisions concerned with the 
final execution of orders and regulations are made by the noncoms 
and the buck privates of the power hierarchy. Such decisions may be 
insignificant from the standpoint of a superior, but they are often 
vitally important for the commoners whose fate they affect. 

The parallel between the lower strata of the apparatus hierarchy 
and the small businessmen of a capitalistic society is obvious. A small 
capitalist has little influence on the conditions of supply, marketing, 
or finance, except when he combines with others of his kind; but 
whether he does so or not, he can usually decide where and what he 
wants to buy and/or produce. In fact, he makes many small decisions 

a. Max Weber drew attention to the fact that under the conditions of supreme bu- 
reaucratic power the mass of the population are all reduced to the level of "the ruled,' 
who see themselves confronted by "a bureaucratically stratified ruling group" that 
actually, and even formally, may occupy "an altogether autocratic position" (Weber, 
WG: 667; cf. 669, 671). 


respecting the small affairs that are his world. Similarly, middle and 
even lower functionaries in hydraulic society are, like the top- 
ranking leadership, part of the power apparatus; and with proper 
grading they, too, enjoy advantages that accrue essentially from 
the unrestricted authority of the regime. 

In terms of income, lower members of the apparatus hierarchy 
may be compared to the employees Of a capitalistic enterprise who 
do not share in the surplus they help to realize. A property-based 
sociology of class would therefore consider them commoners rather 
than members of the upper class. Such an approach, however, over- 
looks the human relations that usually and specifically characterize 
the operations of a bureaucratic order. These operations make the 
lowest representatives of the apparatus state participants in the 
exercise of total power. In contrast to the employees of a commercial 
or industrial enterprise who proceed under the give-and-take con- 
ditions of the market and thus in a formally equal way, even the 
most lowly men of the apparatus proceed on the basis of coercion, 
that is, in a formally unequal way. Their position in the power 
hierarchy provides some of the lowest functionaries with particular 
opportunities for personal enrichment; and it provides all of them 
with a specific sociopolitical status. As representatives of the despotic 
state, even the lowest functionaries arouse in the commoners a 
mixture of suspicion and fear. They therefore occupy a social posi- 
tion which places them, in terms of power, prestige, and sometimes 
also of revenue, outside of, and ambivalently above, the mass of 
the ruled. 

The natives of a conquered country consider the occupying army 
a unit; and they do so knowing full well that the power of the rank- 
and-file soldier is extremely restricted. Similarly, the subjects of a 
hydraulic despotism view the men of the apparatus as a unit, even 
when it is clear that individual members vary immensely in power, 
wealth, and social status. 

2. The Multiple Conditioning of 
Social Subsections 

The ruling class is differentiated from the earliest beginnings of 
hydraulic civilizations. The ruled class is usually undifferentiated in 
simple hydraulic societies. It is always differentiated in semicomplex 
and complex hydraulic societies. 

The subsections of the two classes are differently conditioned. 
Within the ruling class position in the power hierarchy is the primary 
determinant, and wealth, although at times significant, remains 

CHAPTER 8, C 305 

secondary. Within the ruled class types and dimensions of active 
property are the primary determinants o£ social status, whereas 
differences in relation to the government tend, in this apolitical 
world, to play a minor role or no role at all. 


1. The Men of the Apparatus 

a. The Basic Vertical Structure 

The ruling class of hydraulic society is represented first by its active 
core, the men of the apparatus. In virtually all hydraulic countries 
these men are headed by a ruler, who has a personal entourage (his 
court) and who controls and directs his numerous civil and military 
underlings through a corps of ranking officials. This hierarchy, 
which includes the sovereign, the ranking officials, and the under- 
lings, is basic to all Orientally despotic regimes. Horizontal develop- 
ments, which occur under certain conditions, complicate the basic 
vertical structure. 

i. the ruler and the court 

The despot's arbitrary cruelties and his equally arbitrary gener- 
osities form the themes of many records. His arbitrary cruelties in- 
dicate that subject to obvious physical and cultural limitations he 
can make or break anyone if he wants to. His arbitrary generosities 
indicate that, subject to obvious economic limitations, he can spend 
wastefully and without being restricted by any constitutional agency. 
The proverbial glamour of Oriental courts is merely an economic 
expression of the ruler's despotic control over his subjects. 

In his person the ruler combines supreme operational authority 
and the many magic and mythical symbols that express the terrifying 
(and allegedly beneficial) qualities of the power apparatus he heads. 
Because of immaturity, weakness, or incompetence, he may share 
his operational supremacy with an aide: a regent, vizier, chancellor, 
or "prime minister." But the exalted power of these men does 
not usually last long. It rarely affects the symbols of supreme au- 
thority. And it vanishes as soon as the ruler is strong enough to 
realize the autocratic potential inherent in his position. 

The unique importance of the ruler's whims and actions give 
unique importance to individuals who may influence him. In 
addition to the vizier — and sometimes more consequential than he — 
the best situated to do so are the members of the ruler's personal 


entourage: his wives and concubines, his blood relatives and affinals, 
his courtiers, servants, and favorites. Under the conditions of despotic 
autocracy, any one of them may temporarily and irrationally wield 
excessive power. 


In speaking of officials, we refer to persons who are assigned a 
particular type of government task. Among sedentary peoples the 
regular duties involved in such a task tend to be permanently and 
physically located in an "office" or "bureau." And usually the holder 
of such an office keeps a record of his dealings. 

Linguistically, the word "bureaucracy" is a monstrosity. 1 But the 
importance of some of its connotations has made it popular, despite 
the disapproval of the purists. Semantically, a bureaucrat is a person 
who "rules through bureaus." In a specific sense, the term is also 
applied to any official who uses secretarial devices ("red tape") to 
delay action, to make himself important, or to idle on the job. When 
Stalin criticized "bureaucracy," he particularly stressed "bureauc- 
racy and red tape," officials who indulge in "idle chatter about 
'leadership in general,' " "incorrigible bureaucrats and chair- 
warmers." 2 

Certainly bureaucratic chair-warmers can be annoying and harm- 
ful; and even serving and controlled governments are plagued by 
them. But a bureaucracy becomes truly formidable only when its 
offices are the organizational centers of ruthless and total power. For 
this reason Stalin's effort to hide the bureaucratic Frankenstein of 
the Soviet regime behind the semihumorous facade of inefficient 
"chair-warmers" is nothing more than a clumsy attempt at totali- 
tarian myth making. 

The ranking officials include civil and military functionaries of 
recognized status. They do not include the bureaucratic underlings. 
The civil officials resemble their military colleagues in that both are 
in positions of command and able to make limited and intermediate 
decisions, that both are parts of centrally directed bodies, that both 
unconditionally (and usually full time) serve their ruler, and that 
both are government-supported either by salary or by revenue 
derived from state-assigned office lands. 

An army is essentially an instrument of coercion, and as such not 
necessarily a bureaucratic institution. But the management of cen- 
trally directed armies of the Oriental type involves considerable 
organizational planning, which in literate civilizations is usually 
carried out through bureaus. Many officers are both fighters and 

CHAPTER 8, C 307 

administrators; but often the fighting officials are functionally 
separated from the bureaucratic officials (Militarbeamte). In any 
case officers are not feudal knights but government functionaries, 
and as such part of the ranking officialdom. 


The underlings of the bureaucratic hierarchy are either scribes or 
menial aides. The scribes account for the bulk of all the secretarial 
work done at the court, in the central government, and in the 
provincial and local offices. The menial aides act as gate keepers, 
runners, servants, jailers, and, in a semimilitary capacity, as police- 

In all sizable agrobureaucratic despotisms the underlings are 
numerous. During the last period of imperial China about 40,000 
ranking (civil) officials had at their disposal over 1,200,000 clerks 
and over 500,000 runners — that is, a total of more than 1,700,000 
underlings, or something more than forty underlings to one rank- 
ing official. 3 

b. Horizontal Developments 

The bureaucratic network may spread over a large territory. But 
as long as the central government appoints the bulk of the ranking 
officials and directs the provincial bureaus, no special problems of 
horizontal authority arise, even when the regional functionaries, for 
reasons of distance or political expediency, are given a relatively free 
hand in the conduct of their business. 

Max Weber was struck by the relatively loose way in which the 
central government of imperial China controlled the provincial 
bureaucracy; * and indeed, in accordance with the law of diminishing 
administrative returns, regional and local officials were given con- 
siderable freedom of decision in matters of detail. But as Weber 
himself recognized, the central government appointed and trans- 
ferred these officials at will; and it determined the major lines of 
their action. 5 

Of course, from time to time dynastic authority declined; and 
when the inner crisis was serious, the high territorial officials became 
temporarily the semi-autonomous, or even the autonomous, masters 
of the areas they administered. But except for periods of disruption, 
the most distinguished provincial dignitaries were merely prominent 
members of the centrally established and centrally manipulated 
ranking bureaucracy. 



The Persian empire of the Achaemenids differed from the Chinese 
empire both in origin and structure. The unification of China was 
prepared for by centuries of institutional growth; and the core areas 
of Chinese culture were sufficiently populous and strong to make 
their domination over the outlying and colonized regions relatively 
easy. Conversely, the Persians in a single generation extended their 
rule beyond the confines of their homeland to four sizable coun- 
tries, each of which had a well-defined culture: Media (549), Lydia 
(546), Babylonia (538), and Egypt (525). They abolished the ruling 
houses in all four of these regions and in addition changed the politi- 
cal map by carving them up into a number of provinces, each 
governed by a satrap. 6 

The heterogeneity and size of their new acquisitions compelled 
the Persian conquerors to give their satraps unusual freedom in the 
handling of political affairs. A satrap might retain his position for a 
long period; and at times he might be succeeded by his son. 7 More- 
over, he appointed the subsatraps 8 and probably also the local 
officials, who were usually natives. 9 He hired mercenary troops and 
his bodyguard. He commanded militia-like levies raised in his terri- 
tory. 10 He administered the taxes of his province. 11 He maintained 
diplomatic relations with neighboring states. 12 And he might organize 
a military expedition against a neighbor country — usually, however, 
with the permission of the Great King. 13 Surrounded by his court, 
he ruled with kinglike splendor. 1 * This quasiroyal status of the satrap 
was actively encouraged by the Persian sovereign, 15 who apparently 
considered this as good a way as any to maintain his prestige in dis- 
tant regions. 

Nevertheless, in several crucial respects, the Great King exerted 
strict control over his satraps. Definitely and conspicuously, he 
was the master, the satrap the serving man who owed him absolute 
obedience. A central system of communication and intelligence, 16 
inspection by metropolitan officials, 17 and the maintenance of Persian 
garrisons at strategic points 18 prevented the satrap from attaining 
military or fiscal independence. The satrapies were taxed according 
to centrally devised principles and with definite obligations toward 
the capital. "The proceeds of this taxation were forwarded annually 
by the satraps to Susa, where the surplus that remained, after de- 
fraying the annual outgoings, accumulated in the king's treasury 
as a reserve fund." 19 

The Great King considered his satrap not a feudal vassal but a 
top-ranking territorial agent. "The king is the master over all his 

CHAPTER 8, C 309 

subjects and the satrap his representative; they can arbitrarily in- 
terfere everywhere, not only where this is required by the realm's 
interest, but wherever they want." 20 

Thus the Persian empire was "a bureaucratic state" {ein Beamten- 
staat); 21 and the satrap's administrative and military freedom of 
action did not destroy the basic structure of the bureaucratic hier- 
archy of which the satrap was a part. 


A satrap might be a native of the region over which he had juris- 
diction. But this was not typical. Only in Cilicia did the Great King 
permit a member of the former ruling house to become the governor 
of a newly established province. 22 Princes who voluntarily accepted 
Persian sovereignty were generally permitted to continue ruling as 
vassals. Like the satraps and subsatraps, they owed the Great King 
military services and tribute; 23 but they seem to have enjoyed more 
political and cultural freedom than did many other native rulers who 
fell under the sway of powerful hydraulic empires. 

The builders of the Inca empire permitted rulers who had sur- 
rendered voluntarily to hold official position; but these curacas 
were subordinated to Inca governors." Moreover, the shrines of the 
region's highest deities were transferred to Cuzco; and the main 
features of Inca religion were imposed upon the new subjects. 24 
While in some ways perpetuating the appearance of native rule, the 
curacas were, for all practical purposes, integral parts of the imperial 
officialdom. 6 

In Muslim India a number of native "chiefs" or rulers (rais, rajas) 
were also, if somewhat differently, included in the ruling order. A 
raja was permitted to preserve many secondary features of his 
previous power, if he vowed unconditional political (and fiscal) 
submission to his new overlord. Says Moreland: "His [the chief's] 
tenure depended on his loyalty, which meant primarily the punctual 
payment of tribute." 25 The rajas were more or less free to determine 
how, in their regions, the tribute should be raised. 26 In Akbar's 
time the six older provinces, which constituted the core of the em- 
pire, were almost completely administered by the central govern- 

a. Usually their sons were taken as hostages to Cuzco, where they were taught the 
Inca way of life (Rowe, 1946: 272). 

b. They were chiefs of 10,000, 5,000, 1,000, or 500 corviable men. See Rowe, 1946: 
263. The chiefs of 100 were apparently the lowest ranking officials. Like the higher 
functionaries, they might ceremonially participate in communal agriculture; but essen- 
tially they supervised and directed the chiefs of ten, who as foremen worked with 
the peasants (Rowe, 1946: 263, 265). 


ment, whereas the outlying provinces presented a mixed picture, 
some being headed by centrally appointed officials, some by rajas. 27 

The Persian satraps, the Inca curacas, and the rajas of Muslim 
India constitute a series of variants on the scale of political sub- 
ordination. The relations between a satrap or curaca and their 
sovereign were definitely noncontractual; in substance, as well as in 
form, the ruler demanded total submission. The position of certain 
rajas included elements of a contractual arrangement; but these were 
expressed factually more than formally. Only with regard to the 
most loosely attached dependencies did a despotic overlord accept, 
under the cloak of an alliance, a quasicontractual relationship. 

The contrast to feudal patterns of subordination is manifest. 
Under a feudal regime the contractual relationship is essential; and 
it characterizes the core of the feudal order. Under a hydraulic 
despotism relations of total submission characterize the core of the 
bureaucratic system and they also prevail in its horizontal extensions. 
Only in the loosely dependent periphery do quasicontractual (quasi- 
feudal) features make their appearance. 

The sociology of hydraulic despotism recognizes relevant differ- 
ences between an ordinary member of the centrally directed bureauc- 
racy and a satrap (or curaca), and between such dignitaries and a 
raja or a loosely dependent ally. In all cases the determining force 
is the agrarian apparatus state; but the degree of operational de- 
pendency creates significant subdivisions in the edifice of despotic 


Analysts of modern industrial apparatus states are equally con- 
cerned with differences between the officials of the totalitarian 
heartland and the heads of the satellite countries. In these cases too 
it is essential to recognize both the supreme role of the metropolis 
and the structural differentiations that characterize its horizontal 
extensions." It is also essential to recognize the tendency toward 

c. In 1921 Stalin characterized the horizontal gradations in the newly established 
USSR as follows: "The Russian experiment in applying various forms of federation, 
in passing from federation based on Soviet autonomy (the Kirghiz Republic, the Bash- 
kir Republic, the Tatar Republic, the Gortsi, Daghestan) to federation based on treaty 
relations with independent Soviet republics (the Ukraine, Azerbaidjan), and in allow- 
ing intermediate phases (Turkestan, White Russia), has fully proved the significance 
and flexibility of federation as a general form of state government for the Soviet 
republics" (Stalin, S, V: 22). Stalin considered this allegedly voluntary association a 
transitional step to a future and "supreme" unity; and indeed as far as the then 

CHAPTER 8, C gl! 

intensified subjugation in periods of imperialistic growth and co- 
ordination. The quasi-independent ally of yesterday may be the 
dependent ally of today and the satellite, satrap, or run-of-the-mill 
official of tomorrow. 

In hydraulic society this trend has its counterpart in retrogressive 
developments that may ultimately replace one unified despotic 
regime by several such systems. Pharaonic Egypt fell temporarily 
into a number of quasi-independent territories; and post-T'ang 
China was even more seriously dismembered. But in both cases the 
new political units perpetuated despotic methods of statecraft, and 
the term "feudal," which may with a certain poetic licence be 
applied to the relation between the weakened center and its larger 
sub-units, is completely inappropriate when used to designate sub- 
units which are actually nothing but detached and smaller replicas 
of the larger despotic model. 

The control mechanism of the modern apparatus states makes 
separation extremely difficult as long as the despotic metropolis itself 
persists. The defection of Tito's Yugoslavia was made possible by 
exceptional geomilitary circumstances. 28 Manifestly, the horizontal 
extensions of the modern apparatus state are not identical with — 
although they offer instructive parallels to — the territories of satraps, 
rajas, or dependent allies in hydraulic society. 

2. Subclasses Attached to the Men of 
the Apparatus 

The manipulators of the despotic state apparatus are the core, but 
not the whole of the ruling class. A biosocial supplement — blood 
relations and affinals — must be included, and frequently also an 
operational supplement — persons who enjoy a semi-, quasi-, or pre- 
official status. 

a. Attachment Based on Kinship 


Polygamy was a recognized institution in the great majority of all 
hydraulic societies; d and for obvious reasons the sovereign had 

"autonomous" and "independent" republics were concerned, he and his comrades 
worked successfully to bring this about: "This voluntary character of the federation 
must absolutely be preserved in the future, for only such a federation can become a 
transitional form to that supreme unity of the workers of all countries in a single 
universal economic system, the necessity for which is becoming ever more perceptible" 
(ibid.: 23). 
d. Interesting exceptions: Christian Byzantium and Russia. The prevalence of 


unique opportunities to utilize it. His many relatives (by blood or 
marriage) usually enjoyed a distinguished social status, and usually 
they also enjoyed considerable material advantage. If and to what 
extent the despot employed them in the government depended on 
a number of circumstances. But when employed, they had an ex- 
cellent chance to rise to positions of prominence and power. 

In the Inca empire the male descendants of the sovereigns were 
organized in ayllus, whose number increased with the advance of 
the dynasty. The members of these ayllus "formed a useful court 
circle of educated men trained in the imperial ideology, and in- 
terested in its perpetuation. The emperors chose their top ad- 
ministrators from this group when possible." 29 

In certain Chinese dynasties, such as Han, the consort family 
played an enormous political role; and in the conquest dynasty of 
Liao the members of the consort clan, Hsiao, were apparently more 
trusted than those of the imperial Yeh-lii clan. 30 But whether the 
sovereign's blood relatives or his affinals were numerous or not in the 
bureaucratic hierarchy, the members of these two groups were 
generally a distinguished component of the ruling class. 


On a less exalted level, the families of the ranking officials are 
equally significant. Like the relatives of the sovereign, although not 
entirely for the same reasons, the members of what may be called 
the bureaucratic gentry did not necessarily hold office. Some were 
too young, some too old, some inept, some women; and some who 
had the qualifications could not find a government post, first, be- 
cause there were usually more candidates than vacancies and second, 
because some of the vacancies might be filled by outsiders rather 
than by sons of officials. 

The amount and form of family possessions are important differ- 
entiating factors. Mobile passive wealth (gold, jewels, etc.) shrinks 
more quickly than landed property, which, although it is fragmented 
through equal division among the heirs, may during the owner's 
lifetime remain undiminished, if the rents are large enough to 
support him and his family. Thus hydraulic societies with highly 
developed private landownership provide the bureaucratic gentry 
with optimal, if gradually diminishing, opportunities for living on 

monogamy in Byzantium and Russia shows that this form of marriage, despite the 
restrictions it imposed on the rulers, was nevertheless compatible with the main 
political, economic, and social trends of Oriental despotism. 

CHAPTER 8, C 313 

the amassed family wealth. The Chinese saying that a family may rise 
from rags to riches in three generations and go back to rags in the 
next three well describes the trend toward declining wealth that in 
contrast to the feudal gentry characterizes the bureaucratic gentry 
of hydraulic society. Equally important is the speed with which a re- 
turn to government service can reestablish (or increase) family 
wealth. No doubt, if members of an impoverished gentry family 
held office for three generations, the family fortune (and landhold- 
ings) at the end of that time would certainly be large. But often 
one family member who served the government even for a limited 
time was able to restore his family's property. In a Chinese case with 
which I am personally familiar a three-year stint in a county 
magistracy did the trick. 

The political significance of the bureaucratic gentry is indicated 
by the fact that members of this group are frequently invited to 
fulfill auxiliary administrative, judicial, or priestly functions. In 
Pharaonic Egypt remunerative positions in the temple service were 
often given to children of notables. 31 In the judicial assemblies of 
Babylonia some "notables" were officeholders, others had a gentry- 
like status. 12 

For Buddhist India Fick assumes the existence of a "gentry of the 
land" that formed a part of the gahapatis, the "householders." 33 In 
his opinion these householders were neither warriors, Kshatriyas, 
nor Brahmins; 3 * rather they were identical with, or overlapped, a 
"lower-land-owning nobility." 35 Fick's interpretation of the house- 
holders is open to doubt. Dutoit considers them members of the 
third order, the Vaisya. e The texts which Fick has translated show 
clearly that Brahmins could be householders; 36 and this indeed was 

e. Jdtakam, II: 143, n. 1; cf. IV: 541, n. 1. At this period, castes, jati, already existed. 
But the jati, which later increased to several thousand, are not identical with the 
four major varna, the Kshatriyas, Brahmins, Vaisyas, and Siidras. The use of the 
word varna ("color") as a designation for these great divisions, goes back to the period 
covered by the Rigveda — that is, to the days when the Aryans, the persons of "the 
light color," subdued the indigenous Dasyus, the persons of "the dark color" (Rapson, 
1922: 54; see also Renou, 1950: 63). After this period the term varna "denotes 'a social 
order" independently of any actual distinction of colour" (Rapson, 1922: 54) or a 
"class" or "order." Smith (1928: 36), following Shama Sastri, suggests these or "some 
equivalent term." Cf. also Rhys-Davids, 1950: 46. The rules of castes, jati, which most 
prominently regulate diet and marriage, shaped with increasing rigidity the four 
orders, among which, however, only the Brahmins persisted throughout India and 
until modern times: "No four original castes ever existed at any time or place, and 
at the present moment the terms Kshatriya, Vaisya, and Sudra have no exact meaning 
as a classification of existing castes. In northern India the names Vaisya and Sudra are 
not used except in books of disputes about questions of caste precedence. In the 
south all Hindus who are not Brahmans fall under the denomination of Sudra, while 


their regular condition when, after having completed their educa- 
tion, they married and founded a family. 37 

Fick's classification seems valid, to this extent: a householder en- 
joyed no "special privileges," S8 and when he lived essentially off 
his lands he generally belonged to the lower nobility — that is, to a 
segment of the ruling class that was less distinguished than the 
officeholding Kshatriyas, Brahmins, or Vaisyas. But land grants were 
made primarily to secular serving men and to Brahmins; 39 and the 
nonofficiating members of these groups certainly constituted a bu- 
reaucratic or priestly gentry. This was the case whether they held 
land grants hereditarily or for life, or not at all. 40 

In Byzantine Egypt the relatives of prominent functionaries seem 
to have been eager to assume office when opportunity offered. While 
living on their estates, they fulfilled a variety of semi-official functions 
in their locality. 41 

The Inca state took elaborate measures to support meritorious 
dignitaries and other persons of distinction. The lands assigned to 
them were also intended to benefit their descendants. 42 This suggests 
that, as in other hydraulic civilizations, a sizable bureaucratic gentry 
flourished in Inca society. In pre-Conquest Mexico, too, sinecure 
lands were held for long periods of time, not only by relatives of 
the ruling house but also by the families of ranking officials. 43 

In China individuals who achieved social distinction because of 
their bureaucratic family background can be documented as early as 
the Chou period; and at least since T'ang times and with due con- 
sideration for the degree of relationship, the kinsmen of ranking offi- 
cials enjoyed legally established advantages. 44 Thus they constituted 
a bureaucratic gentry within the terms of our definition. 

In a somewhat different way, Western writers have applied the 
term "gentry" to the shen-shih, the sash bearers, a group that over- 
laps but is not identical with the bureaucratic gentry of the present 
inquiry. As far as we know, the designation shen-shih is found only 
in official documents of the last dynasties. The lists of shen-shih in- 
cluded natives of a particular region who were or had been officials, 
and in addition persons who had attained a degree either, and in the 
main, through examination or through purchase, but who, as yet, 
had not held office. 

The examination system appears relatively late in Chinese history; 
and the classing of holders of examination degrees as a social group 

the designations Kshatriya and Vaisya are practically unknown" (Smith, 1928: 35). The 
consolidation, social rise, and persistence of the Brahmins in Hindu and Muslim India 
is a crucial aspect of the long and complicated history of Indian society. 

CHAPTER 8, C 315 

appears even later. But whatever the initial date, the bureaucratic 
orientation of the shen-shih is clear. As noted above, shen-shih status 
was determined not by relations to land but to government office.' 
The top echelons of the shen-shih hierarchy were composed of pres- 
ent or former officials and holders of high degrees who expected to 
be in office soon. Much larger was the number of lower shen-shih, 
who, holding lower degrees, would have a long wait. However, like 
the high-degree holders, who had not as yet entered government serv- 
ice, the members of the lower shen-shih engaged in all kinds of semi- 
official activities, such as the promotion of local public works, local 
defense and security measures, management of relief and welfare en- 
terprises, and the collection of contributions and fees for the govern- 
ment.*' And they were, of course, always ready to accept a government 
position, which, in addition to opening the way to greater political 
and social influence, was incomparably more rewarding materially. 

/. In an analysis of the Chinese gentry published in 1946, H. T. Fei emphasized 
both its proprietary and bureaucratic aspects; but his formulation of the second point 
remains somewhat vague: "Not until one of the family members [of a landlord] enters 
the scholar group and into officialdom is their position in the gentry consolidated" 
(Fei, 1946: 11; italics mine). In 1948 in a book which he wrote prior to joining the 
Communist camp, Fei, defining the gentry, mentions their connections to government 
office before landownership: "The gentry may be returned officials or the relatives of 
officials or simply educated landowners" (Fei, 1953: 32). To fully appreciate Fei's 
statement it should be remembered that he sharply rejected any idea of a self- 
perpetuating landlordism in China. The law of inheritance dissolved even large hold- 
ings; and traditionally the major road to acquiring land was through government 
office (see Fei and Chang, 1945: 302). This implies that the bulk of China's landowners, and 
particularly the large and educated landowners, were bureaucratic landlords — that is, 
typical members of a bureaucratic gentry. 

Eberhard in a recent definition of the Chinese gentry mentions their "landed 
property^ first; he refers to the proprietary aspect again, and prominently, when he 
describes "the gentry class" as comprising "landowners, scholars and politicians in one 
and the same class," normally with "representatives of all three occupations in one 
family" (Eberhard, 1952: 16; cf. 14. italics in original.). Eberhard "feels not qualified 
to write about Egypt, Mesopotamia and India" (ibid.: 35, n. 2); and he does not 
consider Riistow's concepts of the "Hellenistic-Oriental sultanate" and of the bureau- 
cratic state slavery of the later Roman empire (Riistow, OG, II: 169, 187). Lacking 
crucial tools for a comparative study of Oriental government and property, he 
remains unaware of the peculiar character (and strength) of the former and of the 
peculiar character (and weaknesses) of the latter. 

g. Chang (GI, Pt. II). A number of these tasks, such as construction and repair of 
local roads, irrigation canals, and river dikes and the collection of contributions and 
fees for the government, belong to the intermediate type of enterprises, which in 
hydraulic society are sometimes handled by the bureaucracy, sometimes by private 
persons (cf. Wittfogel, 1931: 413 ff., 445 ff.). These private persons are mostly mem- 
bers of the bureaucratic ruling class, and their work assumes a semi-official character 
when it is backed by government authority for the collection of funds and for the 
mobilization of people. 


The average ranking civil official may have derived about 1700 taels 
from his office.^ The average member of the lower shen-shih prob- 
ably had an annual income in the neighborhood of 200 taels from 
"gentry services." 

For some hydraulic societies our evidence on the existence of a 
bureaucratic gentry is merely suggestive; for others it is conclusive. 
But even where documentation is scanty, the presence of privileged 
members of the ruling house and of a similarly, if less conspicuously, 
privileged bureaucratic gentry seems indicated. The ranking officials 
were eager to share the advantages of their positions with their rela- 
tives. And within the range of their power they certainly did so. 


And then there are the relatives of the civil underlings and rank- 
and-file soldiers. About the day-to-day life of this numerous group 
we know little. In the 17th century in China a bureaucratic rack- 
eteer, Li San, lived splendidly because he was able to cash in on his 
own and his father's and grandfather's experience as government 
clerks. 45 His success, although short-lived and exceptional, underlines 
the benefits that intelligent and ambitious relatives of civil under- 
lings could derive from their position. 

The families of professional soldiers constituted a more or less 
analogous group. Some of their problems are indicated in Ham- 
murabi's Code, 46 and a comparative study of the Ptolemaic cleruchs 
and the peasant soldiers of the Byzantine themes would probably re- 
veal similar conditions. 

For the most part, the relatives of these civil and military un- 
derlings were as modestly situated economically as the bulk of the 
artisans and peasants. But politically and socially they shared the 
ambivalent prestige of their serving family members. The social posi- 
tion, which the father, wife, or son of a policeman enjoys in a police 
state, gives some indication of the place occupied by the relatives of 
underlings in an Orientally despotic state. 

h. These estimates have been made on the basis of the data in Chang, GI, Pt. II. 

CHAPTER 8, C 317 

b. Attachment Based on Semi-, Quasi-, or Pre-official Status 

Not all relatives of the men of the apparatus share to the same de- 
gree the social privileges of their officiating kinsmen. Relative close- 
ness to the bureaucratic activists and the peculiarities of the prevail- 
ing kinship system define the beneficiaries' specific position within 
the ruling class. But whatever the variations, this position, other con- 
ditions being equal, derives from the prominence that the apparatus 
state gives its functionaries. 

In a different way this is true also for groups that have a semi-, 
quasi-, or pre-official status. Although they are not properly officials, 
members of such groups work for the government as economic agents 
or they are granted official or quasi-official status because as func- 
tionaries of the dominant religion they magically bolster the security 
of the regime. 


Persons who spend all or most of their time serving the govern- 
ment as economic agents (damkar, setthi) are sometimes included 
among the officials. In this case their status need not be argued. 
Often, however, commercial agents are not so listed; and fiscal agents 
(tax farmers) are rarely if ever considered part of the bureaucratic 
hierarchy. But although these men are denied official rank, they are 
recognized as servants of the government. In this capacity they are 
supported and given authority, sometimes even coercive authority, 
and to compensate them for their services they are granted a fee or 
commission. In Ptolemaic Egypt tax farmers were granted a fee of 5 
per cent, and later 10 per cent; * 7 in Byzantium of one per cent, 2.5 
per cent, or 5 per cent; 48 in Muslim India up to 10 per cent. 49 In 
Ottoman Egypt they had assigned to them, in addition to cash in- 
come, about 10 per cent of all cultivable village land, the so-called 
wasiya. 60 

To be sure, commercial agents and tax farmers might succeed in 
gathering and keeping more than the prescribed quota. But this 
tendency, which was vigorously combated by strong rulers, does not 
distinguish the economic agents from the commercial or fiscal offi- 
cials, who were equally eager to gather and keep more than they 

The commercial and fiscal agents were private entrepreneurs in 


that both used private means and, in some part, privately hired em- 
ployees. But in acting for the government they enjoyed the advan- 
tages of government authority, and when necessary they could mo- 
bilize government personnel to impose their will. The population 
respected and feared them, not as private individuals but as exten- 
sions of government power. 

If these persons were officials or members of the bureaucratic gen- 
try, who sought to increase their wealth through semi-official opera- 
tions, their bureaucratic position was established a priori. In any 
event, the government-based character of their functions made them 
semi-officials and placed them in, although often at the periphery of, 
the ruling class. 


In a previous chapter we examined the methods by which the agro- 
despotic state closely attached to itself the dominant religion and its 
functionaries. In China and in the earlier period of Pharaonic Egypt 
government officials performed many of the major tasks of the domi- 
nant cult. In other Oriental civilizations the government appointed 
the priests of the dominant religion and from the standpoint of ad- 
ministration treated them like secular officials (Staatsbeamte). 51 

The religious functionaries of Islam lived for the most part on 
endowments (waqfs) that were controlled directly or indirectly by 
the government. 52 In this respect they were more closely attached to 
the state than were the Brahmins of Hindu India, who only on occa- 
sion received land grants. In both cases, however, the state enforced 
the sacred law of the dominant religion, which bestowed a quasi- 
official position of authority on the religious functionary. 

To be sure, any religious functionary enjoys a special kind of awe 
among the believers; but his prestige may be weakened or enhanced 
by the over-all setting in which he operates. The priest of a secondary 
and underprivileged creed may have difficulty in asserting his au- 
thority even among his own followers, who are constantly exposed to 
the disparaging value judgments of an unfriendly environment. The 
priest of a dominant creed is not faced with such difficulties. On the 
contrary. The respect of the rulers enhances his prestige; and the 
more so, the stronger the government happens to be. Under hydraulic 
despotism the functionaries of the dominant religion, even where 
they are not appointed officials, enjoy on the social level a quasi- 
official status. 

CHAPTER 8, C 319 


The intricacies of ideology and script, and most hydraulic civiliza- 
tions above the tribal level had a script, tended to make the training 
for office a drawn-out affair, and those who participated in it often 
constituted a special group. If they were accepted in official "col- 
leges" or "universities," they were carefully selected and restricted 
in number. This was the situation in Aztec Mexico and Byzantium, 
under the Mamluks, in Ottoman Turkey, and during certain periods 
of Chinese history such as Han. 

Where the students were educated in temples and/or by priests, 
their training was not specifically bureaucratic but their number 
was similarly restricted. Where competitive examinations were open 
to the public generally, as they were under the two last Chinese dy- 
nasties, the students were numerous, and the lower degree-holders 
many. Exposed to a long and intensive process of indoctrination, the 
students may well have been even more sensitive to the benefits and 
eminence of the bureaucratic life than the office-holders. Bureau- 
cratic class consciousness could be further consolidated if degree- 
holders were permitted to carry out certain semi-official functions. 
The members of the Chinese shen-shih, who held examination de- 
grees but had not yet attained office, are a classical example of a pre- 
bureaucratic group. 


In hydraulic society the rulers rarely manipulated the sacred doc- 
trine, even if they were its high priests. In the Soviet Union the Or- 
thodox Church, although still tolerated, is no longer a dominant 
creed; and when the openly expressed Soviet designs materialize, 
the Church will be replaced altogether by the secular state doctrine. 
The standard bearers of this doctrine are the masters of the apparatus 
state; they — and they alone — interpret and change it. The country's 
top-ranking ideologists are the top-ranking members of the ruling 
bureaucracy; and the great bulk of all professional intellectuals are, 
like them, government officials. 

A few outstanding artists and writers may produce their works 
without holding office. But they follow state directives, they execute 
state orders, they are paid as are high officials; and since they serve 


the state well and without reservation, they enjoy similar preroga- 
tives. For all practical purposes, they have a quasi-official status. 

The difference is important. While in hydraulic society the quasi- 
official ideological (religious) functionaries are many and relatively 
free insofar as doctrine is concerned, in the USSR the quasi-official 
intellectuals are few and their freedom in matters of doctrine is nil. 
The total managerial state is ideocratic. Its rule involves control over 
the society's ideas. It nationalizes the ideology as well as the ideologists. 

c. Subdivided, but Still an Entity 

Our survey reveals that even under the simplest conditions the 
ruling class in hydraulic society is divided into several subsections. 
Under more differentiated conditions it tends to be a fairly complex 
entity. How conscious of the peculiarity and superiority of their class 
position are the members of the various subsections? 

Class consciousness is probably a less general — and certainly a less 
dynamic — factor than Marxism would have us believe. But there 
can be little doubt that the masters of hydraulic society, who en- 
joyed extraordinary privileges of power, revenue, and status, formed 
one of the most class-conscious groups in the history of mankind. 

To be sure, their class-consciousness did not always express itself 
in images which underlined their greatness as ranking officials. The 
serving men of Ottoman Turkey were proud to be the "slaves" of 
their sultan. The glory of the ruling class, as they saw it, rested upon 
its autocratic ruler. The political ideologists of Hindu India stressed 
the prominence of the king as the supreme protector of the dominant 
religion. The glory of the ruling class, as they saw it, rested upon its 
priestly advisors. The Confucian philosophers paid homage to their 
absolute sovereign; but they extolled the gentleman-scholar, who, 
because of his training, was likely to become a gentleman-bureaucrat. 
The glory of the ruling class, as they saw it, rested upon its properly 
educated officials. 

Confucianism presents the sociopolitical aspect of the matter with 
unusual clarity. By designating the gentleman-scholar as chun-tzu, 
Confucius emphasized the political quality of his ideal man. The 
chun-tzu was thoroughly versed in the cultural tradition of the 
hereditary ("noble") officialdom, but his qualifications had an 
essentially political intent. The word chun-tzit originally connoted 
"a ruler," "a man engaged in the business of ruling." After being 
properly trained, the chun-tzu was ready to be "used" as a govern- 
ment official." He was ready to rule the "little men," the mass of the 


The dichotomy between the two groups finds expression in the 
Chinese terms shih and min. The shih are those individuals who, by 
their training in ethical, military, and ceremonial matters, are 
qualified to serve their ruler and who do so when this is possible. 
The min are "the people," who are ruled by the sovereign and the 
officiating members of the shih. 1 The values placed on civil and 
military qualifications have varied over time.' But the glorification of 
the shih endured until the end of the imperial era. 

Whatever the nomenclature, the shih-min distinction operates in 
all hydraulic societies. In all of them the potential and the actual 
rulers are deeply aware of their superiority to, and difference from, 
the mass of the ruled — the commoners, the "people." 


1. Property-based Subsections of Commoners 

Below the rulers spreads the vast world of 'the commoners. Its 
members share a negative. quality: none participates in the affairs of 
the state apparatus. They also share a positive quality: none are 

Chinese tradition distinguishes three main groups of commoners: 
peasants, artisans, and merchants. The sequence reflects the order of 
their appearance on the historical scene; but it is doubtful whether 
this was in the minds of those who listed them so. More likely they 
were concerned with relative economic importance, agriculture being 
the root (pen) and handicraft and commerce the branches (mo) of 
their agrarian civilization. 

i. The Classic of History frequently refers to the officials as shih (Legge, CC, III: 
275, 367, 369, 626), as also do the Odes (ibid., IV: 360, 409, 429 ff., 569). More narrowly 
the term shih connotes lower ranking officials (cf. ibid., I: 401). As persons of proper 
training, the shih are frequently mentioned, particularly in the Confucian writings 
(cf. ibid., I: 168, 274, 276). The ultimate test of their education is revealed in govern- 
ment service (ibid., I: 271 ff., 339). To be sure, the friendship of a shih should be 
sought even when he is net in office (ibid., I: 297). 

Frequently the shih are juxtaposed to the min. The former serve elegantly at the 
royal ancestral temple (ibid., IV: 569) or the court, while the latter look on and admire 
them (ibid., IV: 409 ff.). Taken together, the shih and the min constitute the whole 
population. In periods of unrest both groups suffer (ibid., IV: 560). 

;'. Confucius was primarily interested in the civil qualifications of the shih; and this 
obviously modified an earlier tradition (see Legge, CC, I: passim; cf. Wittfogel, 1935: 
49, n. 3). 

a. The Chinese classification which places the shih before the peasants, artisans, 
and merchants does not recognize a class of persons whose position rests essentially 
on landownership. 


The root and the branches correspond to two basic forms of 
property: immobile and mobile. In our survey of the complexity 
patterns of property we discussed the rise, development, and social 
position of the three just-mentioned groups in considerable detail; x 
and there is no need to recapitulate our conclusions here. However, 
to round out the inquiry, we shall at this point examine the position 
of the most lowly social group: the slaves. Slaves played only a very 
limited role in hydraulic society. Why? 

2. Slaves 

Soil, water, and plants are manipulated with great care by persons 
who profit personally from their labors: peasant members of village 
communities, owner-cultivators, and tenant farmers. But no such 
care can be expected from full slaves — that is, from persons, who, 
in addition to being personally unfree, possess neither family nor 
property. This is true for agrarian conditions generally, and it is 
especially true for areas where the agronomy is largely determined 
by irrigation farming. 

In irrigation-based hydraulic agriculture slave labor was little 
employed. Occasionally when easy access to slave labor suggested its 
use in farming (or in handicraft), such labor remained an auxiliary 
force. To assure the necessary care the slaves were usually given a 
share in what they produced, and at times they could marry. 

The costs of supervision inhibited the use of great numbers of 
slaves in the most typical of all public works in hydraulic society: 
the construction and maintenance of canals, embankments, roads, 
and walls. It was only in spatially restricted enterprises, such as mines 
and quarries, the building of palaces and temples, and the transport 
of bulky objects that slave labor could be easily supervised and 
therefore advantageously employed. 2 

This explains why state slaves are found primarily at the court, 
in government offices, workshops, and mines, and in special types of 
building activities. It explains why privately owned slaves were 
essentially employed domestically and by wealthy persons, who could 
afford the luxury of lavish consumption. 3 It explains why occasional 
attempts to use slaves in subtler tasks compelled their public and 
private masters to provide conspicuous incentives and to replace full 
slavery by semislavery. 

A victorious war might, of course, produce a sizable slave reservoir. 
And while the conquerors of agricultural regions usually hastened 
to assign the bulk of their peasant captives to farming, the occupa- 

CHAPTER 8, D 323 

tion in which they would best profit their new masters, some might 
be kept as government slaves or sold to private persons. 

The Aztecs, who frequently fought their neighbors, had little use 
for slave labor in their communally organized calpulli villages. But 
as sacrifices at the great state ceremonials many captives served the 
purposes of spectacular terror, a major device for keeping the crudely 
coordinated Mexican empire united. 

In ancient Mesopotamia warfare between the independent states 
was also an important source of slaves; and in Babylonia slaves were 
used to some extent in agriculture and handicrafts. But here, too, 
slave labor remained a secondary feature; and usually it was em- 
ployed under conditions of semislavery: the slaves could acquire 
property and marry.* In Pharaonic Egypt slavery seems to have as- 
sumed some importance only in the New Kingdom, when major wars 
and conquest flooded the country with unfree foreign labor. s 

After examining the history of ancient Mesopotamia and Egypt in 
their entirety, Westermann finds that in these civilizations slave 
labor was predominantly domestic; 6 and Meyer, in his evaluation 
of Near Eastern slavery, asserts that "scarcely anywhere in the Orient 
did slavery play a major economic role." 7 Mendelsohn's recent study 
on slavery in the ancient Orient confirms the earlier findings. What- 
ever slave labor was used in agriculture "was of no great weight. On 
the whole, slaves were used primarily in domestic service." 8 

Studies of other Oriental countries reach the same conclusions. 
There were many slaves in India, China, and the Islamic world, 
but in none of these large civilizations did slave labor dominate 
agriculture or handicraft. 6 

Some slaves and freedmen were raised to positions of prominence 
by Oriental despots, and others were given important supervisory 
tasks by private slave owners. But their careers were not representa- 
tive of the conditions of their group. While the domestic slaves of 
hydraulic society in their majority were not chattel slaves, 9 they 
were personally unfree and they remained at the mercy of their 
masters. In the case of the female slaves it was usually taken for 
granted that their masters had access to them. 

b. For India, see C. A. F. Rhys-Davids, 1922: 205; Fick, 1920: 306 ff. Appadorai 
(1936, I: 317 ff.) does not correlate his findings on the use of slaves in late Hindu 
South India with his analysis of agriculture and industry. But his description of 
these two branches of economy imply what Dr. Rhys-Davids has explicitly noted for 
Buddhist India. In both fields slave labor was insignificant (C. A. F. Rhys-Davids, 
1922: 205). For Chinese society in general, see Wittfogel, 1931: 393 ff.; for the Han 
period, see Wilbur, 1943: 174 ft., 195 ff. For Abbassid society, see Mez, 1922: 152 ff.; 
for pre-Mongol Persia, see Spuler, 1952: 439 ff. 


In a society that polarized total authority and total submission in- 
dividuals who lacked every personal freedom were not to be envied. 
Their position was little improved by the fact that in certain hy- 
draulic civilizations and in wealthy families they were, at times, 


Slavery affects the bottom rung of Oriental society, conquest the 
top. Indeed, conquest may change the traditional structure of a 
conquered area so greatly that we are justified in designating the 
institutional result as a conquest society. 1 The sociology of conquest 
has essentially stressed the relation of conquest to the beginnings of 
stratified societies (primary conquest in our terms); and this process, 
although not too fully recorded, certainly deserves attention. But 
conquest may further differentiate already stratified societies (sec- 
ondary conquest in our terms); and this process, which is more fully 
recorded and which generally involves more recent developments, 
deserves particular attention. 

1. Conquest Involving the Formation of 
Stratified Societies (Primary Conquest) 

War between independent political commonwealths is as old as 
human life. But devices for keeping a given population permanently 
subdued developed only when permanent subjection was both re- 
warding and feasible. Was this possibility exploited first and in all 
cases by conquerors? Or did the increasing facilities of production 
first lead to the emergence of a native upper group, a tribal nobility 
or a professional officialdom? 

Lowie, who considers "internal conditions" sufficient "to create 
hereditary or approximately hereditary classes," 2 cautiously evaluates 
the possible range of internal differentiation and conquest by stating 
that the two factors "need not be mutually exclusive." 3 

An essentially endogenous development has been documented in 
a number of cases, 4 but there seems to be no question that in other 
cases conquest created a conspicuous social stratification and very 
often intensified and advanced incipient endogenous differentia- 
tion. Conquest of this kind — primary conquest — apparently occurred 
throughout the hydraulic world, in ancient Greece and Rome, in 
Japan, and in Medieval Europe. It is a general and not a specific 
factor, and therefore it cannot be held responsible for the diverse 

CHAPTER 8, E 325 

patterns of power, property, and class, which characterized these 

2. Conquest Involving the Further Differentiation 
of Stratified Societies (Secondary Conquest) 

Secondary conquest does not always lead to the establishment of a 
conquest society. The bulk of all members of the conquering group 
may remain in their homeland; and their leaders may be satisfied 
to exercise remote control either by placing their own nationals 
directly over those they have subjugated, or by utilizing native 
collaborators, or by establishing strategically placed garrisons. Rule 
by satraps, curacas, or rajas is usually an end product of military 
conquest; and it involves significant horizontal gradations of power. 
But the resulting institutional order is not a conquest society in the 
sense of the present inquiry. 

I speak of a conquest society only when the conquerors take up 
residence in the lands they have seized, when they neither liquidate 
nor expel the native population, and when they are sufficiently 
numerous to establish a cohesive and distinct alien ruling body 
apart from, and above, their new subjects. 

a. For the history of the relation of conquest to the origin of class structure, see 
Riistow, OG, I: 84 ff. The phenomenon has been systematically discussed from a 
sociological standpoint by Gumplowicz (1905: 190 ff., 195 ff.) and Oppenheimer (1919: 
32 ff.), both of whom promoted the thesis that class differentiation is generally initiated 
by conquest. This thesis has been convincingly challenged by the anthropologists 
MacLeod (1924: passim) and Lowie (1927: 33 ff.). Without consideration for their 
arguments, Riistow (OG, I: 66 ff., 74 ff., 95 ff.) accepts in the main the earlier conquest 
thesis; but he admits the possibility of social differentiations resulting from internal and 
peaceful development (OG, I: 88 ff., 90 ft.), and he recognizes that the conquest-created 
societies are diversely structured. Although he suggests that these societies be called 
" 'medieval' or 'feudal' in the widest sense" (OG, I: 79), he notes that the term 
"feudal" in its "political and narrower meaning" fits essentially Medieval Europe (OG, 
I: 312), that in ancient Rome a big-peasant aristocracy formed the dominant class 
(OG, II: 166), and that in Egypt from the dawn of history a planned economy doomed 
the mass of the population to "state slavery" (OG, II: 187). 

In view of this, it is unfortunate that Eberhard, who "accepts the theory of A. Riistow 
of the power factor, which creates feudal societies by superstratification" (Eberhard, 
1952: 3) and who considers Rustow's ideas "the so far most complete theory on the 
origin of feudalism" (ibid.) fails to familiarize his readers with the structural diversity 
of Rustow's feudal societies. Eberhard sees "no principal difference between Oriental 
and Western feudalism" (ibid.: 2). But we have only to confront Eberhard's feudal 
system "based essentially on land which the vassal held as a fief" (ibid.: 1) with the 
Oriental reality and Rustow's concept of Egypt's "spiritual feudalism" with its ruling 
priesthood and planned state slavery (OG, II: 17, 31, 187) to realize the inadequacy 
of Eberhard's view both from the standpoint of institutional facts and from that of his 
Ueged authority, Dr. Riistow. 


Incipient conquest societies have emerged as the result of primary 
conquests. Full-fledged conquest societies appeared in many parts of 
the world and under a variety of circumstances. Their rise was 
inevitably stimulated by the attractiveness of the target country and 
by the military strength and mobility of the conquerors. Agricul- 
tural civilizations (and particularly "wealthy" hydraulic economies) 
were highly desirable objectives; and until modern times powerful 
nomadic tribes (especially pastoralists who could ride and use the 
saddle and stirrup) have been optimally successful in seizing them. 5 

3. Class Modifications in Hydraulic 
Conquest Dynasties 

a. The Chinese Did Not Ahuays Absorb Their Conquerors 

Great and culturally persistent peoples, such as the Chinese, have 
pointed to the speed with which their "barbarian" conquerors 
adopted many features of their way of life. Easy-going generaliza- 
tions from this cultural fact originated the widespread legend that 
the Chinese "always" absorbed their conquerors. However, reality 
contradicts this legend. Instead of relinquishing their privileges of 
power, prestige, and revenue the conquerors invariably sought to 
maintain them by all manner of political, military, and legal devices. 
And where they found it desirable to do so they also preserved 
particular features of their own cultural tradition. 

Comparative analysis shows that none of the four major conquest 
dynasties of China confirms the myth of absorption, not even the last. 
The Manchus had already adopted many Chinese customs prior to 
the conquest; e but in their case, as in the others, basic differences in 
political and social status were maintained to the end. 6 

b. Devices for Preserving the Conquerors' Hegemony 

The reasons for this are easily understood. The "barbarian" con- 
querors depended for many details of civil administration on native 
experts and bureaucrats. But they protected their political, social, 
and economic hegemony by placing their own nationals above the 
indigenous officialdom, by concentrating their tribal soldiers in 
special cadres, camps, ordus (hordes), or banners, by making inter- 
marriage with the subdued population difficult or impossible, and 

b. Under conditions of conquest, cultural change is closely interlinked with political 
change. Our Chinese findings are therefore suggestive for conquest societies in general: 
"full cultural amalgamation obviously occurred only when the disappearance of the 
social divide permitted the cultural divide to disappear also — that is, after the period 
of conquest had come to an end" (Wittfogel, 1949: 15). 

CHAPTER 8, F 327 

by preserving their tribal religion even when, for purposes of 
prestige, the ruler and his lieutenants performed the great indigenous 

The Arab warriors, who were the military mainstay of the 
Umayyad dynasty, lost their social prominence when that dynasty 
collapsed,** just as in China the Ch'i-tan, Jurchen, Mongols, and 
Manchus lost their privileged position when their respective con- 
quest dynasties (Liao, Chin, Yuan, and Ch'ing) came to an end. 

c. Duplications of Class 

Thus conquest societies tend to involve a curious duplication of 
social strata. As a rule, an exogenous upper class (nobility) is super- 
imposed upon a native bureaucracy; and tribal warriors become a 
distinguished stratum of plebeian underlings in the political hier- 
archy. The newly organized banners, camps, or ordus replace the 
former cadre troops, and definitely outrank the native troops which 
the regime may decide to maintain. 


For obvious reasons the representatives of the despotic state are 
significant in any study of class structure; and this not because the 
men of the apparatus form the bulk of the population — which they 
certainly do not — but because state power, more than any other 
factor, shapes the fate of both the members of the ruling class and 
the commoners. This becomes crystal clear when we consider the 
three major types of social antagonisms that arise in hydraulic 
society: antagonisms between the members of different subsections of 
commoners, antagonisms between the commoners and the state, and 
antagonisms between the members of the various subsections of the 
ruling complex. 

1. Social Antagonism and Class Struggle 

Social antagonism is not identical with class struggle. A conflict 
may be considered social when it involves members of different social 
groups and when it arises essentially out of the social position of 
those concerned. But a social conflict which is limited to a few 

c. This, for instance, happened in the case of the Manchus, whose emperors per- 
formed the traditional Chinese sacrifices, while within the privacy of the palace they 
continued to worship their tribal gods (Wittfogel, 1949: 14). 

d. Wellhausen, 1927: 557. The Umayyads did not conquer the Near East, but they 
consolidated the conquests accomplished under the first caliphs. 


persons cannot reasonably be called a class struggle. The term "class" 
connotes a group) — and usually a relatively large group — of socially 
homogeneous individuals; and a social conflict assumes the character 
of a class conflict only when those who participate in it represent a 
recognizable and representative fraction of such a group. 

Class struggle involves mass action. Such a struggle may reach a 
point where it challenges existing social and political conditions. 
Marx, who perhaps more than any other social scientist of the 19th 
century studied classes, stressed this aspect of the matter by saying 
that "every class struggle is a political struggle." * 

2. Paralysis of Class Struggle by Total Power 

The meaning of all this for an understanding of hydraulic society 
is far-reaching. An agrarian despotism which is strong enough to 
prevent independent political organization does not need to tolerate 
mass action as a means of settling social conflicts. The men of the 
apparatus easily control the secular and religious variants of the 
Beggars' Democracy. They are suspicious of all rallies of socially 
dissatisfied persons. And usually they hasten to break up incipient 
mass movements. 

During the middle period of the Ch'ing dynasty, in 1746, some 
Fukienese tenants joined together in requesting an adjustment of 
their rents. Apparently this was nothing but an argument between 
two groups of private persons, yet the local officials quickly inter- 
vened, arrested, and punished the leaders. 2 A subsequent edict 
blamed the provincial officials for the fact that "stupid people 
assemble and violate the law." 3 

A Han discussion of state and private enterprise in the manu- 
facture of salt and iron objected to private businesses that employed 
more than a thousand workers, since such an accumulation of man- 
power might provide opportunities for treacherous action. 4 At the 
close of the imperial period an edict noted emphatically that there 
had "always been a law of this dynasty forbidding the establishment 
of societies and associations of any sort whatsoever." 5 The statement 
is significant both for its hostility to popular associations and for its 
lack of concern for the existing craft and trade guilds. Obviously 
the government did not count these organizations among the politi- 
cally relevant societies and associations. 

Such attitudes precluded political mass action (class struggle) as a 
legitimate form of social protest. And they did so even in the ruling 
class. Conflicts between members of different subsections of this class 
were often politically colored in that they involved antagonistic 

CHAPTER 8, G 329 

claims to power-based privileges; but they rarely led to open and 
political mass action. The history of hydraulic society suggests that 
class struggle, far from being a chronic disease of all mankind, is 
the luxury of multicentered and open societies. 


In simple hydraulic societies peasants constitute almost the whole 
of the "ruled," and they continue to be the most numerous sub- 
section of commoners in semicomplex and complex hydraulic soci- 
eties. How much opportunity is there for social antagonism between 
them and other commoners? 

Poor (and tenant) farmers may clash with rich (landowning and 
well-to-do) farmers, with traders, or with money lenders. However, 
the possibility of such frictions is minimal in the regulated village 
communities that prevail in the majority of all hydraulic societies. 
For in these communities tenancy is either a nonexistent or a mar- 
ginal issue; and the economic differences between the similarly 
situated peasant households are slight. Moreover, the limited eco- 
nomic flexibility of the average community member restricts the ex- 
tent to which he may deal — and clash — with nonpeasant commoners: 
artisans, traders, and/or money lenders. a 

a. W. C. Smith in his article "Lower-class Uprisings in the Mughal Empire" says 
nothing about social conflicts of this inter-commoner type. Several times he mentions 
"landlords" as involved in class struggles with peasants. But in one instance he only 
surmises the existence of such persons (1946: 28); in others he uses the word "land- 
lords" as an equivalent for zamlndars (ibid.: 27, 30). Until the 18th century the 
zamlndars were essentially tributary rajas (Moreland, 1929: 279); and the "nobility" 
which, according to Smith, seized "approximately one-third of the country's agricul- 
tural produce" did so "in the form of what is called 'taxes' or 'revenue' " (1946: 23). 
That is, these "nobles" were actually government functionaries who lived on govern- 
ment revenue. This pattern is altogether different from the system of land tenure 
of feudal Europe; and it is regrettable that Smith, who was aware of this (1946a: 
308), nevertheless designated the Indian conditions as "feudalism" (ibid.). 

Peasants apparently participated in rebellions of various kinds, but those that can 
be clearly recognized as involving secular issues seem mostly to have arisen from 
fiscal conflicts. As may be expected in a country dominated by rulers of an alien 
creed, religious conflicts frequently merged with secular ones; and in many cases the 
former probably gave voice or increased intensity to the latter (see Smith, 1946: 27 ff.). 
But we have no reason to doubt that certain conflicts were genuinely — or primarily — 
religious. In 1672 members of a small sect clashed with the authorities, defeated the 
local police and several contingents of regular troops, and temporarily controlled the 
city of Narnawl. Smith, who views this event as a "desperate class struggle" (ibid.: 29), 
fails to mention any secular issue which would justify such a classification. 

And then there were struggles that essentially concerned national or territorial 
issues. The Pathan rebellion, which Smith designates as "perhaps the most formidable 


Rural conflicts increased as private landownership increased. In 
Tsarist Russia large peasant uprisings flared up in the 18th century 
when the pomeshchiki became the owners of their former service 
land and when the peasants, encouraged by rumors of all kinds, 
hoped to become the owners of the land they were tilling. 1 The 
reform of the pomeshchik land in 1762 was followed by serious 
peasant disturbances, 2 which reached their climax in the great re- 
bellion led by Pugachev (1772— 75).® 

Conflicts arising out of the usurious lending of grain or money and 
out of oppressive tenancy are well documented for Ptolemaic and 
Roman Egypt, for traditional China, and, of course, for many hy- 
draulic societies in transition. 

Recent studies have often concentrated on these property-based 
conflicts and as a consequence have paid little attention to the 
extraordinary forces of bureaucratic power and property that under- 
lie and complicate the tensions between various groups of wealthy 
and poor commoners. But however much such studies have misunder- 
stood the character of hydraulic society, they provide us with valu- 
able data on conflicts arising from property; and they relieve us of 
the need to repeat here what their authors have industriously, if one- 
sidedly, said on this subject. 

The rise of private property and enterprise in handicraft and 
commerce created conditions that resulted in social conflicts of 
many kinds among urban commoners. In Medieval Europe such con- 
flicts were fought out with great vigor. Not infrequently the social 
movements assumed the proportions of a mass (and class) struggle, 
which in some towns compelled the merchants to share political 
leadership with the artisans and which in others assured the hegem- 
ony of the craft guilds. 4 

The contrast with the hydraulic world is striking. Although the 
guilds of hydraulic society have a much longer history than their 
Western counterparts, they rarely, if ever, engaged in militant and 
political activities of comparable scope. 6 

people's movement" of the Mogul period was the prolonged and pathetic endeavor of 
proud border tribesmen to resist "the attempt to impose . . . [on them] the rule of 
the Mughal State" (ibid.: 33, 34). And in the district of Kishtwar it was obviously a 
semi-independent group of local rulers that combated the Moguls' infringement upon 
them. The protagonists of the Kishtwar rebellion, local zamlndars, defended the cause 
of their prince, who eventually was reinstated. The fact that the "lower classes" also 
"fought and suffered" and that the ryots and inhabitants of nearby Kashmir "com- 
plained" about the harshness of the Mogul commander (ibid.: 27) is scarcely a 
reason for including this affair among the "lower-class uprisings" of the period. 

b. Cf. above, Chap. 4. The Karimi merchants of Mamluk Egypt accumulated great 
fortunes in the international spice trade and as bankers; and their commerce with 

CHAPTER 8, H 331 


The disproportion between the intensity of social antagonism and 
the frequency of class struggle becomes particularly striking when 
we view the relations between the two main classes of hydraulic 
society: the "people" and the men of the apparatus. In the normal 
course of events the commoners suffer periodically from the demands 
made on them by representatives of the despotic state. Generally 
those who are oppressed or exploited do not dare to resist openly; 
and frequently they do not even dare to resist covertly. The Oriental 
subject's proverbial eagerness to avoid any contact at all with the 
feared organs of government underlines his acceptance of defeat in 
a contest that he never dares to enter. 

Avoidance, however, is not always possible. The commoner may 
not lay his complaints before a judge or magistrate; but often he 
must render services and usually he must pay a tax. He may bitterly 
resent both demands, and being unable to protect himself by con- 
stitutional means, he may feign compliance. But behind this facade 
he will combat the men of the apparatus with all the weapons of in- 
direct and passive resistance at his command. 

When he performs his corvee iabor, he will work as slowly as the 
overseer's control (or the stick or whip) permits. 1 When he renders 
his tax, he will seek to conceal certain of his assets. And not in- 
frequently he will hand over his quota only after being severely 
beaten. Writers in Pharaonic Egypt have satirized this aspect of the 
battle of the land tax; 2 and a 19th-century account shows the 
Egyptian peasant's attitude in these matters to be unchanged: "All 
the fellaheen are proud of the stripes they receive for withholding 
their contributions, and are often heard to boast of the number of 
blows which were inflicted upon them before they would give up 
their money." ° 

When taxation becomes unusually burdensome, the peasant may 
reduce his cultivated acreage, 3 and when the heavy demands con- 
tinue, he may become a fiscal fugitive, 6 abandoning his fields alto- 

such countries as Yemen may occasionally have influenced the foreign policy of the 
Mamluk government, which derived great revenues from it. But their economic 
importance notwithstanding, the KarimI merchants failed to attain an independent 
political position comparable to that of the guild merchants of feudal Europe. See 
Fischel, 1937: 72 ff., 76 ft., 80 ff.; cf. Becker, IS, I: 186, 214. 

a. Lane, 1898: 143 ff. Lane adds: "Ammianus Marcellinus gives precisely the same 
character to the Egyptians of his time." Ammianus lived in the 4th century a.d. 

b. The founder of the Mogul Dynasty, Babur, was infuriated by the Indian peasants. 


gether. He may wander in despair, look for work elsewhere, or turn 
bandit or rebel. 

As stated above, open conflicts between peasants and government 
were rare where land tenure was regulated; and even in imperial 
China they assumed major proportions mainly during periods of 
disintegration which initiated the collapse of a dynasty. 

Conflicts between urban commoners (or groups of commoners) and 
the government occurred in a different context. They too frequently 
centered around tax issues; but the administrative (and garrison) 
character of most hydraulic towns generally prevented the discon- 
tented townspeople from resorting to armed rebellion. The in- 
dividual merchants or artisans defended themselves as well as they 
could against restrictive regulations and fiscal exploitation; and 
the guilds of craftsmen and traders, headed by government-appointed 
or government-supervised functionaries, not infrequently appealed 
to the authorities for the adjustment of excessive demands At times 
artisans ceased to work and merchants closed their shops; 4 and 
occasionally a crowd might start a riot. d Government officials who 

who, typical fiscal refugees, hid in the woods and "trusting to their inaccessible situation, 
often continue in a state of revolt, refusing to pay their taxes" (see Babur, 1931: 208). 

c. Chinese historiography relates many such cases (cf. Wittfogel and Feng, 1949: 420). 
An incident that occurred in the Ming dynasty is illuminating in several respects. 
Between 1436 and 1448 a tenant, Teng Mao-ch'i, became a person of influence among 
his fellow villagers, whom he is said to have gotten to "work for him." His prestige 
was greatly enhanced through his leadership in a movement that urged the tenants not 
to make the customary gift to their landlords when they paid their rent. The land- 
lords approached the local magistrate, and it may well be that some of them were 
members of the court or the officialdom, since in Ming days these groups were extremely 
successful in appropriating peasant land. In any case, the magistrate dispatched armed 
forces; but Teng defeated them with a rebel army, which eventually numbered several 
tens of thousands. Soon his power spread over twenty counties, and he received further 
assistance from people who escaped from the "unbearable" oppression of a "greedy 
and cruel" official. Subsequent developments revealed excessive corvee labor as a major 
reason for their discontent. After several military successes the rebels were defeated; 
and Teng, together with some of his followers, was beheaded (Ming Shih 165.53-0). 
An episode in the middle of the fight characterizes both the strength of the govern- 
ment and the limited objectives of the revolt. Negotiating with a courageous official, 
the rebels are reported to have asked only that their lives be spared and that they 
be exempted from labor service "for three years." If these conditions were granted, 
they would lay down their arms, and again be "good people" (Ming Shih 165.5b). At 
the close of the dynasty the government would probably have been more ready to com- 
promise and the rebels less eager to submit. During the last phase of the Ming period 
rebels appeared everywhere; and the many local conflicts were merged in the final 
battle for the overthrow of the dynasty. 

d. For Mamluk Egypt cf. Poliak, 1934: 267 ff. The members of the Indian sect who, 
in 1672, started an uprising are said to have been "goldsmiths, carpenters, sweepers, 
tanners, and other ignoble beings" (commoners?). Some apparently engaged in agri- 

CHAPTER 8, H 333 

were charged with maintaining the rulers' rationality minimum, were 
expected to heed such warnings. And indeed they often did. But 
they were most ready to do so where private, and not state, business 
was involved, 8 and any sporadic moderation on their part did not 
keep them from exerting their authority fully and coercively in 
matters of consequence — for instance, over artisans and laborers who 
rendered corvee labor a or over particular persons whose wealth they 
wished to syphon off. 

In the great majority of all cases the artisan or merchant who 
aroused the greed of a ranking official or underling maneuvered 
prudently. Whenever he could he paid his way out of his impasse. 
Obviously, an accommodating lie or a well-placed bribe are not 
exactly weapons in a war of liberation. And the unending small 
conflicts between the bureaucratic hunter and his petty-bourgeois or 
capitalist game made it unmistakably clear that in this chase the 
urban commoners might survive, but they could not win. 

Traditional Chinese statecraft gave more leeway to private prop- 
erty than did the absolutist regimes of most other hydraulic civiliza- 
tions; but under its shadow capitalist enterprise was as cagey as 
elsewhere. An edict of the short-lived reform government of 1898 
puts the blame for this fact on the officials, particularly — and some- 
what hypocritically — on the underlings. When a firm is in diffi- 

culture (Elliot and Dowson, 1877: 185, 294). Smith (1946: 29) suggests that the urban 
sectarians were workers or poor traders: "petty traders and workers, either property- 
less proletariat or men with a very small professional 'property." His second source 
speaks of trade "on a small scale" or, according to another translation, "their trade 
is on a small capital" (ibid.: 29 ff.). In Muslim India, as elsewhere, propertyless persons 
certainly participated in urban riots; but in this case the cited data point to artisans 
who owned their means of production rather than to proletarian elements. 

Another insurrection of the period is even farther removed from being proletarian. 
According to Smith (ibid.: 25 ff.), the town of Patna was seized in 1610 "by a 
proletarian mob" whose leader impersonated "the popular hero Khusraw." After the 
success of this coup "numbers of the lower-class aligned themselves with him. These 
proletarians even organized a minor army from amongst themselves, which they were 
foolish enough to send out against the upper-class army advancing under the irate 
governor." This account is greatly at variance with the facts as given in Smith's own 
sources. The popular hero Khusraw was the emperor's oldest son, who was kept 
prisoner after he had made an armed attempt to seize the throne (Jahangir, 1909: 
56-68). Khusraw had based his rebellion primarily on the support of members of 
the imperial army (ibid.: 52, 55, 58); and temporarily his chances of success had been 
considerable (ibid.: 58). It is therefore not surprising that the impostor found adherents 
among "a number of foot- and horsemen." These soldiers — and no "proletarian mob" — 
seized Patna and its fort (ibid.: 174); and there is no specific evidence that the "wretched 
creatures," who later joined the rebellion (ibid.: 174), were "proletarians" either. 
Jahangir applies the term "wretch" indiscriminately to rebels, including persons of 
the highest political and social status (ibid.: 55, 65, 123). 


culties, "the demands and extortions of the yamen underlings are 
invariably so great and exorbitant that merchants become dis- 
couraged and dare not venture further afield into trade enterprise, 
thereby causing trade to stand immovable." 7 

The covert conflicts between state slaves and their bureaucratic 
masters were numerous and, generally speaking, unnoticed. Like 
the domestic slaves of private owners, the unfortunate bondsmen of 
the government tried to ease their fate by cunning and well-camou- 
flaged devices; and like them also they were employed essentially as 
single individuals or in small groups and with little opportunity to 
revolt en masse. 

The slave war that started in southern Mesopotamia in 869 drew 
its initial strength from the unusually large number of slaves em- 
ployed by unusually large private enterprises 8 in the production of 
salt, east of Basra. The magnitude of these enterprises made them an 
ideal breeding ground for mass action. The revolt, which lasted some 
fourteen years, owed much of its temporary success to the fact that 
during these years the Abbassid state was shaken by civil wars 
between certain generals and high territorial officials and between 
both and the caliphate. 9 


Except for the peasant uprisings that occasionally, and particularly 
in hydraulic societies with strongly developed private landownership, 
challenged the authority of the officialdom, only the social conflicts 
within the ruling class had a definitely political quality. The military 
rebellions of dissenting members of the ruling family or of ambitious 
generals or governors against a weak monarch usually involved con- 
flicts between persons of different grades and positions within the 
power hierarchy. But they occurred only sporadically and at long 
intervals; and when they did, they tended to evolve quickly into 
military tests of strength between two or more independent terri- 
tories or regions. 

Much more frequent, and much more difficult to discern, are the 
undercover conflicts that arose between ranking officials and bureau- 
cratic underlings, between various groups of ranking officials, be- 
tween officials and the bureaucratic gentry, and between ranking 
officials and the despot and his personal entourage, the court. These 
conflicts were usually concerned with political power or influence, 
and while most of them affected only a few individuals, some of them 
involved the privileges of larger groups, subsections or strata within 
the bureaucratic order. But although such conflicts might touch the 

CHAPTER 8, I 335 

interests of a considerable number of persons, they lacked the or- 
ganized cohesiveness which characterized the great social movements 
of the ancient, medieval, and modern West. 

i. Ranking Officials versus Underlings 

Broadly speaking, the ranking officials determine the operations 
of their secretarial and menial underlings. But often an administra- 
tive (or fiscal or police) problem can be solved to the advantage of 
either the ranking officials or their underlings. Ambivalent situations 
of this kind are inherent in all organizations whose functions are 
vertically divided. But in the hydraulic setting these situations were 
particularly consequential because the actions of the apparatchiki 
were not checked by effective outside forces and because those in- 
volved in the conflicts disposed over the resources of a uniquely 
powerful state apparatus. 

The ranking officials, as well as the underlings, aimed at a maxi- 
mum of control over details of procedure and personnel, partly for 
the sake of power and partly for the sake of increasing their share of 
the government revenue. Status was no major issue, although the 
underlings, by increasing their power, also increased their social 
prestige. A critical examination of the Chinese government under 
the Manchus suggests that the underlings for some time arrogated 
to themselves something like 30 per cent of the government revenue. 1 
Since this estimate was made by a member of the ranking official- 
dom, 2 it may be too high, but it indicates the dimension of the 
economic problem involved in the day-to-day struggle between the 
gentlemen-functionaries and their plebeian aides. 

In this struggle the underlings could and did draw advantage from 
their intimate knowledge of local affairs, their familiarity with the 
know-how of the office, and their physical control over the ultimate 
execution of all administrative work. The officials could and did 
draw advantage from the various methods of supervision, from con- 
trol over the hiring and firing of the staff personnel, and in serious 
cases from the power to invoke all manner of punishment. 

An official Chinese statement of 1899 reveals how in the tug-of-war 
between the ranking officials and the underlings certain functionaries 
might become dependent on strategically placed scribes: "In all 
matters of promotion, transfer, appointment, merit or demerit, or 
of taxes and legal decisions, provincial officials sought to gain favors 
by bribing clerks in the various Boards. And officials who were 
charged with the delivery of revenue, or copper, or dye materials to 
the central government were especially harried by their demands. 


From the day they reported deliveries to the time they were given 
the receipts the clerks found many reasons for making extortions. 
The sums asked reached hundreds and thousands of taels. This 
was known as the 'Board of Expenses' and was collected with little 
effort at concealment." 3 

The runners exerted their power on a different plane and, of 
course, with different methods. They controlled access to govern- 
ment buildings; they arrested people and guarded the jails. Thus 
they could alleviate a prisoner's lot or make it more miserable; they 
could regulate the force of a flogging; they could claim resistance 
to arrest.* The power and possible material benefits inherent in 
these situations are manifest. 

The ranking officials, who wished to maintain their control over 
the numerous and well-entrenched host of underlings, brought into 
play all the administrative and disciplinary means with which they 
were invested. The functionaries of Ch'ing China attempted to limit 
the duration of the underlings' employment. But while such con- 
trol strengthened the hold of the ranking officials over the lower 
functionaries, the costs in skill and experience could be considerable. 

Underlings who abused their power to the manifest detriment of 
the government were to be severely punished. This aspect of the 
matter has been clearly defined in the Arthashastra, in the dynastic 
regulations of China, and in other manuals of agrodespotic state- 
craft. For scribes and runners who were dishonest or resorted to 
extortion the last Code of imperial China established penalties 
ranging from fines to permanent exile and execution by strangula- 
tion. The attached cases show that the higher officials did not hesitate 
to strike when they saw fit. 5 

In the struggle between the officials and the underlings the latter 
could never be completely subdued. But neither could they upset 
the structure of the bureaucratic apparatus, which enabled the rank- 
ing officials to emerge over time, not as total victors but as the 
holders of superior legal, administrative, and economic authority. 

2. Bureaucratic Competition 

a. Patterns of Competition Different in Different Societies 

Competition in the market is only one of many forms of com- 
petition. And hydraulic and feudal society differ from capitalism not 
because in them competition is absent, but because it is differently 

In the medieval world of the West serfdom reduced competition 
in most villages to insignificance, whereas the feudal knights openly 

CHAPTER 8, I gsj7 

and violently competed with their fellows for land and glory. The 
guilds severely restricted competition in the crafts but not in large- 
scale and international trade. 6 

The regulated villages of Oriental society had little opportunity for 
economic rivalries. In traditional China the advance of private 
peasant landownership encouraged competition in economic affairs, 
without, to be sure, making Chinese agriculture capitalist. In all 
types of hydraulic society the members of the ruling class competed 
for power, prestige, and income; and this is true not only for the 
ranking officials but also, and with proper modifications, for the 
bureaucratic underlings. 

Within the capitalist system we find competition on both the 
employer and employee levels. But while the expansion of this 
system increases the quality of goods and the number of persons 
involved, it reduces the number of competing and bargaining ele- 
ments through the rise of corporations and labor unions. In addi- 
tion, legal controls tend to restrict the methods of the competitive 
struggle, which generally is more violent in the early than in the 
later phases of capitalist, economy. 

The difference between the three types of competition appears 
also as a difference in their results. The medieval knight who makes 
a crucial mistake while competing with his fellows (on the battle 
field) may forfeit his life, but his property and honor usually remain 
untouched. The modern businessman who makes a crucial mistake 
while competing with his fellows (on the market), may lose his 
property, but his honor is rarely besmirched, and he certainly will 
not forfeit his life. The official of an agrarian despotism who makes 
a crucial mistake while competing with his fellows (in a bureaucratic 
or court intrigue) is likely to lose his honor, his property, and his 
life. Where power is fragmented and balanced, punishment for a 
crucial mistake is limited. Under conditions of total power, it is 

b. Bureaucratic Competition in Hydraulic Society 

All bureaucratic organizations have certain technical features in 
common; and some methods of intrabureaucratic competition ap- 
pear universally in serving, controlled, and ruling bureaucracies. 
However, this makes it all the more imperative to recognize, behind 
the familiar trees, the peculiarity of the woods of which they form 
a part. a 

a. Universals of warfare appear in the military enterprises of feudal Europe as well 
as in the hydraulic and modern industrial societies. But no one concerned with in- 
stitutional specification will, for this reason, deny the peculiarities of organization and 
procedure that distinguish the three patterns. 


The functionaries of Occidental absolutism are closest to those 
of Oriental absolutism insofar as the chance for a meteoric rise or 
fall is concerned; but under Western absolutism there are non- 
bureaucratic roads to social prominence. And the government officials 
of an open modern society have legally established rights which 
guarantee that the loser in an intrabureaucratic fight need suffer 
nothing more than the frustration of not being promoted. 

Under the conditions of total power bureaucratic life is as com- 
petitive as it is dangerous. A statistical study of the officials of the 
first long-lasting dynasty of imperial China, Han, shows that among 
those whose careers can be traced in some detail 7 about 2 1 per cent 
at one time or another were imprisoned for derelictions during their 
official career, and about 35 per cent died a violent death outside the 
battlefield. More than 1 2 per cent were murdered or died after torture 
in prison, 14 per cent were executed, and 9 per cent committed 

3. Civil versus Military Officials 

Bureaucratic competition occurs not only between members of 
the same office or administrative unit but also between members of 
different branches of the state apparatus. Among these branches, the 
army, for obvious reasons, poses special problems. 

a. The Autocrat and the Army 

The army, as the compact machine of institutionalized coercion, 
plays a different role in different phases of hydraulic society. During 
the formative period the supreme military leader is also apt to 
control the new political economy, since his organizational and 
disciplinary position prepares him uniquely to head the emerging 
agromanagerial apparatus. Once established, the over-all political 
apparatus tends to prevail over the various branches, because the 
heads of the former through their control over personnel and com- 

b. A study of 19th-century China suggests that at the close of the imperial period 
the career of an official was still beset with many dangers, although the character of 
these dangers had changed in several respects. On the basis of the Tung-hua-lu, 
Dr. Hellmut Wilhelm assumes that between 1821 and 1895 "almost every high official 
was punished at least once during his career." Extremely severe punishments (execu- 
tion, banishment, enslavement, corporal punishment, or imprisonment) were imposed 
in about 22 per cent of all cases brought to the emperor's attention, dismissal in 42 
per cent, and lighter punishments (reprimands, fines, and/or demotion) in the remain- 
ing cases. The survey, which considers both Manchu and Chinese officials, was made 
under Dr. Wilhelm's direction, at the University of Washington, Seattle, by Cecil 
Cody, Robert Crawford, Chen-i Wang, and Lincoln Wong. 


munication penetrate all segments, which, no matter what their 
economic weight or coercive potential may be, remain compart- 
mentalized and thus strategically inferior to the coordinating center. 
To elaborate upon our previously established thesis we may say: it 
is not the technical specialist or the hydraulic manager or the head of 
the police, or the commander of the army, but the master of the all- 
pervasive political apparatus who maintains supreme power over the 
compartmentalized technicians, managers, police chiefs, and generals. 
Only during periods of political disintegration and civil war will a 
vigorous general seize control of the entire country or a number of 
generals simultaneously in separated territories become military and 
political leaders: bureaucratic warlords. 

The agromanagerial despot is usually very much aware of the 
power potential inhering in the armed forces; and he therefore takes 
every precaution to keep them subdued. He is the supreme master 
of the military, first because he makes the crucial decisions concern- 
ing its organization, its personnel, and (often also) its supply, and 
second because he heads the centralized apparatus of communica- 
tions and intelligence. 

Similar sociostrategic advantages favor the political masters of 
modern industrial apparatus states. They largely explain why, in the 
ig3o's Stalin was able to liquidate the discontented heads of the 
Soviet army and two subsequent chiefs of the GPU, and why, in 
1944, the National Socialist center prevailed over the generals who 
sought to overthrow Hitler. 

b. Civil versus Military Officials 

The military functionaries, like their civil colleagues, are part of 
the over-all officialdom, and not infrequently the duties of the two 
groups overlap. When essential civil and military tasks are con- 
currently executed by the same higher officials (a governor, a 
satrap, etc.), conflicts between military and civil functionaries occur 
only on lower levels of authority. Often, however, the two spheres 
of action are represented by two distinct groups; and then such 
conflicts appear in the top echelons of the hierarchy. 

Outside of periods of formation, decay, and crisis, military leaders 
in the hydraulic world have a chance to establish positions of promi- 
nence under several conditions: (1) in all areas — core and margin — 
which, being situated between strong neighbors, for international 
reasons require strong protection; (2) in marginal areas, because the 
lesser importance of the managerial bureaucracy increases the weight 
of the army; and (3) in conquest societies, in which the army is an 


essential factor not only for the establishment of the regime but for 
its perpetuation. 

A number of the states of Buddhist India fall into the first category, 
Middle and Late Byzantium and post-Mongol Russia in the second, 
and many conquest societies of the Old and the New World in the 

The struggle between the civil and military officialdom can be 
clearly observed in several hydraulic civilizations. In Pharaonic 
Egypt functionaries who specialized in the military arts proper 
("front" officers) were during prolonged periods subordinated to 
military administrators — that is, to officials who kept the military 
records and who organized supply and equipment. 8 But in another 
context the front officers might successfully counterbalance members 
of the civil administration. The king placed some of them in im- 
portant government positions, where, as socially inferior homines 
novi, they could be relied upon to uphold his interests against the 
ambitions of the ranking civil officials. 9 

Under the Mamluks the military officers, who were exclusively 
Mamluks, remained apart from, and above, the native bureaucracy. 
They could — and did — expropriate, imprison, and execute civil offi- 
cials ivhen they felt the latter were overstepping their authority. 10 

During the last period of the Roman republic successful generals 
rose to the top of the political hierarchy; and under the empire the 
army played a dominant, although varying, role for centuries. 11 

Ostrogorsky considers "the struggle between the competing forces 
of the metropolitan civil aristocracy and the provincial military 
aristocracy" the basic trend of Byzantine society. 12 The meaning of 
this statement becomes clear when we remember that the Byzantine 
civil aristocracy was a Beamtenadel, an aristocracy of officials, 13 and 
that-both groups competed within the framework of a Beamtenstaat, 
a bureaucratic state, which was "constantly swelling and which, as 
the ruling stratum, made ever-greater demands." 14 

The intragovernmental struggles in T'ang China and in com- 
parable periods in the history of other hydraulic civilizations were 
largely struggles between the civil and military branches of the 
ranking officialdom. 

4. The Bureaucratic Activists versus the 
Bureaucratic Gentry 

Conflicts between the officiating functionaries and members of 
the bureaucratic gentry resemble the intrabureaucratic struggles in 
that they, too, are frequently linked with the intrigues and machina- 

CHAPTER 8, I 341 

tions of competing court cliques. However, they have important 
peculiarities of their own. The active bureaucrats wield power; the 
members of the bureaucratic gentry exert influence. The officiating 
executives have excellent opportunities to accumulate wealth; the 
bureaucratic rentiers have fair opportunities to preserve, at least 
during their lifetime, what wealth they have. These differences in 
position go far to explain the conflicts that occur between members 
of the two groups. 

If the individuals concerned are of the same rank, then, other 
things being equal, power will prevail over influence and the 
executive over the rentier. Not infrequently, however, a local official 
of minor rank may find himself in opposition to members of the 
gentry, who are able to prevail because they belong to a bureau- 
cratically powerful family. The study of powerful families in hydrau- 
lic society 15 reveals the decisive role that power plays in this society 
in determining status, influence, and revenue. 

A gentry-bureaucracy conflict may involve only a single member 
of the gentry, a person, let us say, who seeks through influence to 
decrease his fiscal obligations or increase his landholding. Occa- 
sionally it may involve all the members of a local gentry who are 
seeking to shape local politics according to their interests. Members 
of the gentry may stress (and actually represent) the ruler's ration- 
ality maximum; and they may dramatize their intentions by getting 
commoners to demonstrate against the local officials. To support 
their interests at the local level they may even appeal to top-ranking 
members of the hierarchy. 

In the province of Anhui, after the T'ai-p'ing Rebellion, members 
of the gentry, together with other landowners, were temporarily 
able "to cheat the government yearly of a large proportion of income 
from land revenue." The local officials accepted this condition for a 
time because they feared that an insistence on full tax payments 
would cause the people, "incited by the landed gentry," to rebel 
against the newly arrived magistrate. Eventually, however, some 
undaunted members of the bureaucracy suggested the restoration of 
the destroyed cadaster in order to reestablish government control 
over the revenue. 16 

Conversely, several members of the gentry of a certain region in 
the province of Chekiang were dissatisfied with the district magistrate 
because of his "extortions." They complained to his superiors, 
requesting his demotion. 17 

An imperial decree of April 14, 1890, deplored "the common 
practice among the provincial gentry and literati of mixing them- 
selves up in matters of public business, and sometimes even bringing 


pressure to bear on the authorities." The former justified their 
actions by stating that they promoted the public good. However, 
according to the official view these actions were "in reality designed 
for selfish purposes." ls The publication of the edict shows that the 
local officials, who were temporarily at a disadvantage, eventually 
and through the support of the central government, prevailed over 
the gentry. 

In periods of political decay the gentry asserts itself in various 
ways, but officials of a strong regime usually insist that it meet 
their demands. This last was the case in Early and Middle Byzantium 
and in 19th-century Russia, where the negotiations concerning the 
emancipation of the serfs revealed the relative strength of the 
bureaucratic and the proprietary (gentry) wings of the ruling no- 
bility. Theoretically speaking, the (bureaucratic) landowners, or the 
absolutist state and its functionaries, or the peasants might have be- 
come the chief beneficiaries of the Emancipation of 1861. Actually, 
the government one-sidedly determined that the "Editing Commis- 
sion" was to be "composed of officers of the various departments 
which had to do with peasant affairs, together with a number of 
experienced landowners." 19 Thus the terms of the Emancipation 
were "settled by discussion in the bureaucratic field"; 20 and both 
the bureaucratic landowners and the officials presented their respec- 
tive arguments, which were based "not upon any ideal, but upon 
the recognition of the needs of the landowners or of the State." 21 
The bureaucratic quality of the nobles' landed interests was ex- 
pressed in the person of the man who finally headed the Commission, 
Count Panin. Panin owned enormous estates and twenty-one thou- 
sand serfs, but he also had a prominent role in the juridical affairs 
of the government. Pressured by the Tsar and his aides, Panin 
readily subordinated the proprietary aspirations of the nobility to 
its bureaucratic interests. 22 

The relations between the bureaucratic activists and the rentier- 
like bureaucratic gentry recall patterns of conflict occurring in the 
big corporations of modern industrial society. Shareholders of a 
company, who are not among its officers, have the right at the annual 
meeting to comment on, or question, company policy. But such 
casual and optative participation is far from effective control. Satis- 
fied with their dividends, the majority of the shareholders are willing 
to leave the actual management to the executive officers. These 
functionaries exert supreme power over decision making and per- 
sonnel; and even if originally they possessed little stock, they have 
incomparably greater opportunities for improving their material 
position than do the shareholders. 23 

CHAPTER 8, I 343 

In contrast to the corporation shareholders, who have the right to 
assemble, to rally public opinion, and to resort to legal action, the 
members of the hydraulic gentry, even when they owned consider- 
able amounts of land, could not organize or gather freely. The power 
of organized action was restricted to the men in office, who, control- 
ling the bulk of the country's surplus and monopolizing coercive pow- 
er, had no difficulty in stressing the bureaucratic against the propri- 
etary interests of the ruling class. And they did so, even when, as in the 
case of Count Panin, they were both officials and big landowners. 

Thus the conflicts between the bureaucratic gentry and the rank- 
ing officials once more thrust into sharp relief the unique power 
position enjoyed in hydraulic society by the men of the state ap- 

5. Conflicts between the Autocrat and Other 
Members of the Ruling Class 

The autocrat has been likened to the life-giving sun, to fierce 
animals, and to the merciless forces of lightning, storm, and flood. 
To his subjects he is indeed all these, and those among them who 
act in his name are eager both to execute his will and to influence it. 
But the master of a tool is also its servant. The autocrat depends 
operationally upon the persons who implement his orders. The his- 
tory of Oriental courts records endless attempts to influence the auto- 
crat and equally endless attempts by the ruler to prevail over all 
personal and impersonal (bureaucratic) forces. The resulting con- 
flicts are many. Contemplating the autocrat's antagonistic relations 
with his relatives on the one hand and with his ranking officials on 
the other, we can distinguish several types of conflict and also several 
major devices that the antagonists employ to further their respective 

a. The Autocrat versus His Relatives 


The ruler's relatives (who they are depends upon the prevailing 
patterns of kinship) are ever-ready to use their socially privileged 
position for political purposes. To name a successor outside the 
established tradition or to replace a ruler in his lifetime is a risky 
venture; but attempts to do so have frequently been made and not 
always without success. 

Serious problems may arise even when the established tradition 
is upheld. How does an autocrat control his crown prince? How 


does he control his kinsmen? The Han emperors granted them much 
property but little power. Such a policy cannot eliminate all con- 
flicts, but it will restrict them greatly, and to the decided advantage 
of the autocrat. 


The ruler's affinals are an equally ambivalent asset. They attain 
political prominence because one of their female members is his 
wife. They thus have a vested interest in the person of the ruler, who 
on his part may trust them more than his blood relatives. The Han 
rulers almost invariably kept their blood relatives out of office, but 
many members of the empress' family were given high positions in 
the bureaucracy. The Liao emperors were less discriminatory, but 
they, too, often turned to their in-laws when key political positions 
were to be filled. 24 Of course, such a policy has its dangers. Affinals 
who wield great power may reduce the ruler to a figurehead during 
his lifetime. Or after his death they may install a child as his suc- 
cessor and then reign in his stead. During a great part of the Liao 
dynasty the empire was ruled by empress dowagers. 25 

How does an autocrat control his affinals? The limiting of politi- 
cal eunuchism tends to decrease the influence of the ruler's wives, 
and measures designed to protect the heir apparent also have 
obvious advantages. The Toba ruler went to extremes: He killed 
his wife after she bore him an heir. 28 But such radical means were 
rarely invoked. More often, instead of killing the mother of his 
son (or sons), the ruler filled his harem with slave girls. Their rela- 
tives were usually persons of lowly status, and although some among 
them might rise to high station, they were much less of a threat, as 
a group, than noble and well-established consort families. Several 
Chinese emperors were the sons of former "singing girls," 27 and 
the majority of all caliphs c and Turkish sultans had former slave 
girls as mothers. 28 

The problems raised by the blood relatives contrast sharply with 
those raised by the affinals. With regard to the former the ruler 
could narrow the basis of hostility; with reward to the latter he 
could, under optimal circumstances, remove it alt